From 9ed3c5cc09c55d2fffa67b59d9d21e3b44d7653e Mon Sep 17 00:00:00 2001 From: Jedrzej Kosinski Date: Sat, 20 Sep 2025 18:10:39 -0700 Subject: [PATCH 001/114] [Reviving #5709] Add strength input to Differential Diffusion (#9957) * Update nodes_differential_diffusion.py * Update nodes_differential_diffusion.py * Make strength optional to avoid validation errors when loading old workflows, adjust step --------- Co-authored-by: ThereforeGames --- comfy_extras/nodes_differential_diffusion.py | 33 +++++++++++++++----- 1 file changed, 26 insertions(+), 7 deletions(-) diff --git a/comfy_extras/nodes_differential_diffusion.py b/comfy_extras/nodes_differential_diffusion.py index 98dbbf102..255ac420d 100644 --- a/comfy_extras/nodes_differential_diffusion.py +++ b/comfy_extras/nodes_differential_diffusion.py @@ -5,19 +5,30 @@ import torch class DifferentialDiffusion(): @classmethod def INPUT_TYPES(s): - return {"required": {"model": ("MODEL", ), - }} + return { + "required": { + "model": ("MODEL", ), + }, + "optional": { + "strength": ("FLOAT", { + "default": 1.0, + "min": 0.0, + "max": 1.0, + "step": 0.01, + }), + } + } RETURN_TYPES = ("MODEL",) FUNCTION = "apply" CATEGORY = "_for_testing" INIT = False - def apply(self, model): + def apply(self, model, strength=1.0): model = model.clone() - model.set_model_denoise_mask_function(self.forward) - return (model,) + model.set_model_denoise_mask_function(lambda *args, **kwargs: self.forward(*args, **kwargs, strength=strength)) + return (model, ) - def forward(self, sigma: torch.Tensor, denoise_mask: torch.Tensor, extra_options: dict): + def forward(self, sigma: torch.Tensor, denoise_mask: torch.Tensor, extra_options: dict, strength: float): model = extra_options["model"] step_sigmas = extra_options["sigmas"] sigma_to = model.inner_model.model_sampling.sigma_min @@ -31,7 +42,15 @@ class DifferentialDiffusion(): threshold = (current_ts - ts_to) / (ts_from - ts_to) - return (denoise_mask >= threshold).to(denoise_mask.dtype) + # Generate the binary mask based on the threshold + binary_mask = (denoise_mask >= threshold).to(denoise_mask.dtype) + + # Blend binary mask with the original denoise_mask using strength + if strength and strength < 1: + blended_mask = strength * binary_mask + (1 - strength) * denoise_mask + return blended_mask + else: + return binary_mask NODE_CLASS_MAPPINGS = { From 7be2b49b6b3430783555bc6bc8fcb3f46d5392e7 Mon Sep 17 00:00:00 2001 From: Kohaku-Blueleaf <59680068+KohakuBlueleaf@users.noreply.github.com> Date: Sun, 21 Sep 2025 09:24:48 +0800 Subject: [PATCH 002/114] Fix LoRA Trainer bugs with FP8 models. (#9854) * Fix adapter weight init * Fix fp8 model training * Avoid inference tensor --- comfy/ops.py | 13 +++++++------ comfy/weight_adapter/loha.py | 8 ++++---- comfy/weight_adapter/lokr.py | 4 ++-- comfy/weight_adapter/lora.py | 4 ++-- comfy/weight_adapter/oft.py | 2 +- comfy_extras/nodes_train.py | 18 ++++++++++++++++++ 6 files changed, 34 insertions(+), 15 deletions(-) diff --git a/comfy/ops.py b/comfy/ops.py index 55e958adb..9d7dedd37 100644 --- a/comfy/ops.py +++ b/comfy/ops.py @@ -365,12 +365,13 @@ class fp8_ops(manual_cast): return None def forward_comfy_cast_weights(self, input): - try: - out = fp8_linear(self, input) - if out is not None: - return out - except Exception as e: - logging.info("Exception during fp8 op: {}".format(e)) + if not self.training: + try: + out = fp8_linear(self, input) + if out is not None: + return out + except Exception as e: + logging.info("Exception during fp8 op: {}".format(e)) weight, bias = cast_bias_weight(self, input) return torch.nn.functional.linear(input, weight, bias) diff --git a/comfy/weight_adapter/loha.py b/comfy/weight_adapter/loha.py index 55c97a3af..0abb2d403 100644 --- a/comfy/weight_adapter/loha.py +++ b/comfy/weight_adapter/loha.py @@ -130,12 +130,12 @@ class LoHaAdapter(WeightAdapterBase): def create_train(cls, weight, rank=1, alpha=1.0): out_dim = weight.shape[0] in_dim = weight.shape[1:].numel() - mat1 = torch.empty(out_dim, rank, device=weight.device, dtype=weight.dtype) - mat2 = torch.empty(rank, in_dim, device=weight.device, dtype=weight.dtype) + mat1 = torch.empty(out_dim, rank, device=weight.device, dtype=torch.float32) + mat2 = torch.empty(rank, in_dim, device=weight.device, dtype=torch.float32) torch.nn.init.normal_(mat1, 0.1) torch.nn.init.constant_(mat2, 0.0) - mat3 = torch.empty(out_dim, rank, device=weight.device, dtype=weight.dtype) - mat4 = torch.empty(rank, in_dim, device=weight.device, dtype=weight.dtype) + mat3 = torch.empty(out_dim, rank, device=weight.device, dtype=torch.float32) + mat4 = torch.empty(rank, in_dim, device=weight.device, dtype=torch.float32) torch.nn.init.normal_(mat3, 0.1) torch.nn.init.normal_(mat4, 0.01) return LohaDiff( diff --git a/comfy/weight_adapter/lokr.py b/comfy/weight_adapter/lokr.py index 563c835f5..9b2aff2d7 100644 --- a/comfy/weight_adapter/lokr.py +++ b/comfy/weight_adapter/lokr.py @@ -89,8 +89,8 @@ class LoKrAdapter(WeightAdapterBase): in_dim = weight.shape[1:].numel() out1, out2 = factorization(out_dim, rank) in1, in2 = factorization(in_dim, rank) - mat1 = torch.empty(out1, in1, device=weight.device, dtype=weight.dtype) - mat2 = torch.empty(out2, in2, device=weight.device, dtype=weight.dtype) + mat1 = torch.empty(out1, in1, device=weight.device, dtype=torch.float32) + mat2 = torch.empty(out2, in2, device=weight.device, dtype=torch.float32) torch.nn.init.kaiming_uniform_(mat2, a=5**0.5) torch.nn.init.constant_(mat1, 0.0) return LokrDiff( diff --git a/comfy/weight_adapter/lora.py b/comfy/weight_adapter/lora.py index 47aa17d13..4db004e50 100644 --- a/comfy/weight_adapter/lora.py +++ b/comfy/weight_adapter/lora.py @@ -66,8 +66,8 @@ class LoRAAdapter(WeightAdapterBase): def create_train(cls, weight, rank=1, alpha=1.0): out_dim = weight.shape[0] in_dim = weight.shape[1:].numel() - mat1 = torch.empty(out_dim, rank, device=weight.device, dtype=weight.dtype) - mat2 = torch.empty(rank, in_dim, device=weight.device, dtype=weight.dtype) + mat1 = torch.empty(out_dim, rank, device=weight.device, dtype=torch.float32) + mat2 = torch.empty(rank, in_dim, device=weight.device, dtype=torch.float32) torch.nn.init.kaiming_uniform_(mat1, a=5**0.5) torch.nn.init.constant_(mat2, 0.0) return LoraDiff( diff --git a/comfy/weight_adapter/oft.py b/comfy/weight_adapter/oft.py index 9d4982083..c0aab9635 100644 --- a/comfy/weight_adapter/oft.py +++ b/comfy/weight_adapter/oft.py @@ -68,7 +68,7 @@ class OFTAdapter(WeightAdapterBase): def create_train(cls, weight, rank=1, alpha=1.0): out_dim = weight.shape[0] block_size, block_num = factorization(out_dim, rank) - block = torch.zeros(block_num, block_size, block_size, device=weight.device, dtype=weight.dtype) + block = torch.zeros(block_num, block_size, block_size, device=weight.device, dtype=torch.float32) return OFTDiff( (block, None, alpha, None) ) diff --git a/comfy_extras/nodes_train.py b/comfy_extras/nodes_train.py index c3aaaee9b..9e6ec6780 100644 --- a/comfy_extras/nodes_train.py +++ b/comfy_extras/nodes_train.py @@ -38,6 +38,23 @@ def make_batch_extra_option_dict(d, indicies, full_size=None): return new_dict +def process_cond_list(d, prefix=""): + if hasattr(d, "__iter__") and not hasattr(d, "items"): + for index, item in enumerate(d): + process_cond_list(item, f"{prefix}.{index}") + return d + elif hasattr(d, "items"): + for k, v in list(d.items()): + if isinstance(v, dict): + process_cond_list(v, f"{prefix}.{k}") + elif isinstance(v, torch.Tensor): + d[k] = v.clone() + elif isinstance(v, (list, tuple)): + for index, item in enumerate(v): + process_cond_list(item, f"{prefix}.{k}.{index}") + return d + + class TrainSampler(comfy.samplers.Sampler): def __init__(self, loss_fn, optimizer, loss_callback=None, batch_size=1, grad_acc=1, total_steps=1, seed=0, training_dtype=torch.bfloat16): self.loss_fn = loss_fn @@ -50,6 +67,7 @@ class TrainSampler(comfy.samplers.Sampler): self.training_dtype = training_dtype def sample(self, model_wrap, sigmas, extra_args, callback, noise, latent_image=None, denoise_mask=None, disable_pbar=False): + model_wrap.conds = process_cond_list(model_wrap.conds) cond = model_wrap.conds["positive"] dataset_size = sigmas.size(0) torch.cuda.empty_cache() From d1d9eb94b1096c9b3f963bf152bd6b9cd330c3a4 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sat, 20 Sep 2025 19:09:35 -0700 Subject: [PATCH 003/114] Lower wan memory estimation value a bit. (#9964) Previous pr reduced the peak memory requirement. --- comfy/supported_models.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/comfy/supported_models.py b/comfy/supported_models.py index 1fbb6aef4..4064bdae1 100644 --- a/comfy/supported_models.py +++ b/comfy/supported_models.py @@ -995,7 +995,7 @@ class WAN21_T2V(supported_models_base.BASE): unet_extra_config = {} latent_format = latent_formats.Wan21 - memory_usage_factor = 1.0 + memory_usage_factor = 0.9 supported_inference_dtypes = [torch.float16, torch.bfloat16, torch.float32] @@ -1004,7 +1004,7 @@ class WAN21_T2V(supported_models_base.BASE): def __init__(self, unet_config): super().__init__(unet_config) - self.memory_usage_factor = self.unet_config.get("dim", 2000) / 2000 + self.memory_usage_factor = self.unet_config.get("dim", 2000) / 2222 def get_model(self, state_dict, prefix="", device=None): out = model_base.WAN21(self, device=device) From 27bc181c49249f11da2d8a14f84f3bdb58a0615f Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sun, 21 Sep 2025 16:48:31 -0700 Subject: [PATCH 004/114] Set some wan nodes as no longer experimental. (#9976) --- comfy_extras/nodes_wan.py | 4 ---- 1 file changed, 4 deletions(-) diff --git a/comfy_extras/nodes_wan.py b/comfy_extras/nodes_wan.py index 9cca6fb2e..b1e9babb5 100644 --- a/comfy_extras/nodes_wan.py +++ b/comfy_extras/nodes_wan.py @@ -287,7 +287,6 @@ class WanVaceToVideo(io.ComfyNode): return io.Schema( node_id="WanVaceToVideo", category="conditioning/video_models", - is_experimental=True, inputs=[ io.Conditioning.Input("positive"), io.Conditioning.Input("negative"), @@ -375,7 +374,6 @@ class TrimVideoLatent(io.ComfyNode): return io.Schema( node_id="TrimVideoLatent", category="latent/video", - is_experimental=True, inputs=[ io.Latent.Input("samples"), io.Int.Input("trim_amount", default=0, min=0, max=99999), @@ -969,7 +967,6 @@ class WanSoundImageToVideo(io.ComfyNode): io.Conditioning.Output(display_name="negative"), io.Latent.Output(display_name="latent"), ], - is_experimental=True, ) @classmethod @@ -1000,7 +997,6 @@ class WanSoundImageToVideoExtend(io.ComfyNode): io.Conditioning.Output(display_name="negative"), io.Latent.Output(display_name="latent"), ], - is_experimental=True, ) @classmethod From 1fee8827cb8160c85d96c375413ac590311525dc Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 22 Sep 2025 13:49:48 -0700 Subject: [PATCH 005/114] Support for qwen edit plus model. Use the new TextEncodeQwenImageEditPlus. (#9986) --- comfy/text_encoders/llama.py | 16 +++++++---- comfy_extras/nodes_qwen.py | 55 ++++++++++++++++++++++++++++++++++++ 2 files changed, 65 insertions(+), 6 deletions(-) diff --git a/comfy/text_encoders/llama.py b/comfy/text_encoders/llama.py index 5e11956b5..c5a48ba9f 100644 --- a/comfy/text_encoders/llama.py +++ b/comfy/text_encoders/llama.py @@ -400,21 +400,25 @@ class Qwen25_7BVLI(BaseLlama, torch.nn.Module): def forward(self, x, attention_mask=None, embeds=None, num_tokens=None, intermediate_output=None, final_layer_norm_intermediate=True, dtype=None, embeds_info=[]): grid = None + position_ids = None + offset = 0 for e in embeds_info: if e.get("type") == "image": grid = e.get("extra", None) - position_ids = torch.zeros((3, embeds.shape[1]), device=embeds.device) start = e.get("index") - position_ids[:, :start] = torch.arange(0, start, device=embeds.device) + if position_ids is None: + position_ids = torch.zeros((3, embeds.shape[1]), device=embeds.device) + position_ids[:, :start] = torch.arange(0, start, device=embeds.device) end = e.get("size") + start len_max = int(grid.max()) // 2 start_next = len_max + start - position_ids[:, end:] = torch.arange(start_next, start_next + (embeds.shape[1] - end), device=embeds.device) - position_ids[0, start:end] = start + position_ids[:, end:] = torch.arange(start_next + offset, start_next + (embeds.shape[1] - end) + offset, device=embeds.device) + position_ids[0, start:end] = start + offset max_d = int(grid[0][1]) // 2 - position_ids[1, start:end] = torch.arange(start, start + max_d, device=embeds.device).unsqueeze(1).repeat(1, math.ceil((end - start) / max_d)).flatten(0)[:end - start] + position_ids[1, start:end] = torch.arange(start + offset, start + max_d + offset, device=embeds.device).unsqueeze(1).repeat(1, math.ceil((end - start) / max_d)).flatten(0)[:end - start] max_d = int(grid[0][2]) // 2 - position_ids[2, start:end] = torch.arange(start, start + max_d, device=embeds.device).unsqueeze(0).repeat(math.ceil((end - start) / max_d), 1).flatten(0)[:end - start] + position_ids[2, start:end] = torch.arange(start + offset, start + max_d + offset, device=embeds.device).unsqueeze(0).repeat(math.ceil((end - start) / max_d), 1).flatten(0)[:end - start] + offset += len_max - (end - start) if grid is None: position_ids = None diff --git a/comfy_extras/nodes_qwen.py b/comfy_extras/nodes_qwen.py index fff89556f..49747dc7a 100644 --- a/comfy_extras/nodes_qwen.py +++ b/comfy_extras/nodes_qwen.py @@ -43,6 +43,61 @@ class TextEncodeQwenImageEdit: return (conditioning, ) +class TextEncodeQwenImageEditPlus: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "clip": ("CLIP", ), + "prompt": ("STRING", {"multiline": True, "dynamicPrompts": True}), + }, + "optional": {"vae": ("VAE", ), + "image1": ("IMAGE", ), + "image2": ("IMAGE", ), + "image3": ("IMAGE", ), + }} + + RETURN_TYPES = ("CONDITIONING",) + FUNCTION = "encode" + + CATEGORY = "advanced/conditioning" + + def encode(self, clip, prompt, vae=None, image1=None, image2=None, image3=None): + ref_latents = [] + images = [image1, image2, image3] + images_vl = [] + llama_template = "<|im_start|>system\nDescribe the key features of the input image (color, shape, size, texture, objects, background), then explain how the user's text instruction should alter or modify the image. Generate a new image that meets the user's requirements while maintaining consistency with the original input where appropriate.<|im_end|>\n<|im_start|>user\n{}<|im_end|>\n<|im_start|>assistant\n" + image_prompt = "" + + for i, image in enumerate(images): + if image is not None: + samples = image.movedim(-1, 1) + total = int(384 * 384) + + scale_by = math.sqrt(total / (samples.shape[3] * samples.shape[2])) + width = round(samples.shape[3] * scale_by) + height = round(samples.shape[2] * scale_by) + + s = comfy.utils.common_upscale(samples, width, height, "area", "disabled") + images_vl.append(s.movedim(1, -1)) + if vae is not None: + total = int(1024 * 1024) + scale_by = math.sqrt(total / (samples.shape[3] * samples.shape[2])) + width = round(samples.shape[3] * scale_by / 8.0) * 8 + height = round(samples.shape[2] * scale_by / 8.0) * 8 + + s = comfy.utils.common_upscale(samples, width, height, "area", "disabled") + ref_latents.append(vae.encode(s.movedim(1, -1)[:, :, :, :3])) + + image_prompt += "Picture {}: <|vision_start|><|image_pad|><|vision_end|>".format(i + 1) + + tokens = clip.tokenize(image_prompt + prompt, images=images_vl, llama_template=llama_template) + conditioning = clip.encode_from_tokens_scheduled(tokens) + if len(ref_latents) > 0: + conditioning = node_helpers.conditioning_set_values(conditioning, {"reference_latents": ref_latents}, append=True) + return (conditioning, ) + + NODE_CLASS_MAPPINGS = { "TextEncodeQwenImageEdit": TextEncodeQwenImageEdit, + "TextEncodeQwenImageEditPlus": TextEncodeQwenImageEditPlus, } From e3206351b07852f2127a56abd898ee77f7f4c25f Mon Sep 17 00:00:00 2001 From: Christian Byrne Date: Mon, 22 Sep 2025 14:12:32 -0700 Subject: [PATCH 006/114] add offset param (#9977) --- server.py | 9 ++- tests/execution/test_execution.py | 105 +++++++++++++++++++++++++++++- 2 files changed, 112 insertions(+), 2 deletions(-) diff --git a/server.py b/server.py index 43816a8cd..603677397 100644 --- a/server.py +++ b/server.py @@ -645,7 +645,14 @@ class PromptServer(): max_items = request.rel_url.query.get("max_items", None) if max_items is not None: max_items = int(max_items) - return web.json_response(self.prompt_queue.get_history(max_items=max_items)) + + offset = request.rel_url.query.get("offset", None) + if offset is not None: + offset = int(offset) + else: + offset = -1 + + return web.json_response(self.prompt_queue.get_history(max_items=max_items, offset=offset)) @routes.get("/history/{prompt_id}") async def get_history_prompt_id(request): diff --git a/tests/execution/test_execution.py b/tests/execution/test_execution.py index 8ea05fdd8..ef73ad9fd 100644 --- a/tests/execution/test_execution.py +++ b/tests/execution/test_execution.py @@ -84,6 +84,21 @@ class ComfyClient: with urllib.request.urlopen("http://{}/history/{}".format(self.server_address, prompt_id)) as response: return json.loads(response.read()) + def get_all_history(self, max_items=None, offset=None): + url = "http://{}/history".format(self.server_address) + params = {} + if max_items is not None: + params["max_items"] = max_items + if offset is not None: + params["offset"] = offset + + if params: + url_values = urllib.parse.urlencode(params) + url = "{}?{}".format(url, url_values) + + with urllib.request.urlopen(url) as response: + return json.loads(response.read()) + def set_test_name(self, name): self.test_name = name @@ -498,7 +513,6 @@ class TestExecution: assert len(images1) == 1, "Should have 1 image" assert len(images2) == 1, "Should have 1 image" - # This tests that only constant outputs are used in the call to `IS_CHANGED` def test_is_changed_with_outputs(self, client: ComfyClient, builder: GraphBuilder): g = builder @@ -762,3 +776,92 @@ class TestExecution: except urllib.error.HTTPError: pass # Expected behavior + def _create_history_item(self, client, builder): + g = GraphBuilder(prefix="offset_test") + input_node = g.node( + "StubImage", content="BLACK", height=32, width=32, batch_size=1 + ) + g.node("SaveImage", images=input_node.out(0)) + return client.run(g) + + def test_offset_returns_different_items_than_beginning_of_history( + self, client: ComfyClient, builder: GraphBuilder + ): + """Test that offset skips items at the beginning""" + for _ in range(5): + self._create_history_item(client, builder) + + first_two = client.get_all_history(max_items=2, offset=0) + next_two = client.get_all_history(max_items=2, offset=2) + + assert set(first_two.keys()).isdisjoint( + set(next_two.keys()) + ), "Offset should skip initial items" + + def test_offset_beyond_history_length_returns_empty( + self, client: ComfyClient, builder: GraphBuilder + ): + """Test offset larger than total history returns empty result""" + self._create_history_item(client, builder) + + result = client.get_all_history(offset=100) + assert len(result) == 0, "Large offset should return no items" + + def test_offset_at_exact_history_length_returns_empty( + self, client: ComfyClient, builder: GraphBuilder + ): + """Test offset equal to history length returns empty""" + for _ in range(3): + self._create_history_item(client, builder) + + all_history = client.get_all_history() + result = client.get_all_history(offset=len(all_history)) + assert len(result) == 0, "Offset at history length should return empty" + + def test_offset_zero_equals_no_offset_parameter( + self, client: ComfyClient, builder: GraphBuilder + ): + """Test offset=0 behaves same as omitting offset""" + self._create_history_item(client, builder) + + with_zero = client.get_all_history(offset=0) + without_offset = client.get_all_history() + + assert with_zero == without_offset, "offset=0 should equal no offset" + + def test_offset_without_max_items_skips_from_beginning( + self, client: ComfyClient, builder: GraphBuilder + ): + """Test offset alone (no max_items) returns remaining items""" + for _ in range(4): + self._create_history_item(client, builder) + + all_items = client.get_all_history() + offset_items = client.get_all_history(offset=2) + + assert ( + len(offset_items) == len(all_items) - 2 + ), "Offset should skip specified number of items" + + def test_offset_with_max_items_returns_correct_window( + self, client: ComfyClient, builder: GraphBuilder + ): + """Test offset + max_items returns correct slice of history""" + for _ in range(6): + self._create_history_item(client, builder) + + window = client.get_all_history(max_items=2, offset=1) + assert len(window) <= 2, "Should respect max_items limit" + + def test_offset_near_end_returns_remaining_items_only( + self, client: ComfyClient, builder: GraphBuilder + ): + """Test offset near end of history returns only remaining items""" + for _ in range(3): + self._create_history_item(client, builder) + + all_history = client.get_all_history() + # Offset to near the end + result = client.get_all_history(max_items=5, offset=len(all_history) - 1) + + assert len(result) <= 1, "Should return at most 1 item when offset is near end" From 8a5ac527e60fcd48ec228d309d49ab28ac79def8 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 22 Sep 2025 14:26:58 -0700 Subject: [PATCH 007/114] Fix bug with WanAnimateToVideo node. (#9988) --- comfy_extras/nodes_wan.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/comfy_extras/nodes_wan.py b/comfy_extras/nodes_wan.py index b1e9babb5..6c16a2673 100644 --- a/comfy_extras/nodes_wan.py +++ b/comfy_extras/nodes_wan.py @@ -1210,7 +1210,7 @@ class WanAnimateToVideo(io.ComfyNode): background_video = background_video[video_frame_offset:] background_video = comfy.utils.common_upscale(background_video[:length].movedim(-1, 1), width, height, "area", "center").movedim(1, -1) if background_video.shape[0] > ref_images_num: - image[ref_images_num:background_video.shape[0] - ref_images_num] = background_video[ref_images_num:] + image[ref_images_num:background_video.shape[0]] = background_video[ref_images_num:] mask_refmotion = torch.ones((1, 1, latent_length * 4, concat_latent_image.shape[-2], concat_latent_image.shape[-1]), device=mask.device, dtype=mask.dtype) if continue_motion is not None: From 707b2638ecd82360c0a67e1d86cc4fdeae218d03 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 22 Sep 2025 14:34:33 -0700 Subject: [PATCH 008/114] Fix bug with WanAnimateToVideo. (#9990) --- comfy_extras/nodes_wan.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/comfy_extras/nodes_wan.py b/comfy_extras/nodes_wan.py index 6c16a2673..b0bd471bf 100644 --- a/comfy_extras/nodes_wan.py +++ b/comfy_extras/nodes_wan.py @@ -1229,7 +1229,7 @@ class WanAnimateToVideo(io.ComfyNode): character_mask = character_mask.unsqueeze(1) character_mask = comfy.utils.common_upscale(character_mask[:, :, :length], concat_latent_image.shape[-1], concat_latent_image.shape[-2], "nearest-exact", "center") if character_mask.shape[2] > ref_images_num: - mask_refmotion[:, :, ref_images_num:character_mask.shape[2] + ref_images_num] = character_mask[:, :, ref_images_num:] + mask_refmotion[:, :, ref_images_num:character_mask.shape[2]] = character_mask[:, :, ref_images_num:] concat_latent_image = torch.cat((concat_latent_image, vae.encode(image[:, :, :, :3])), dim=2) From 145b0e4f79b5d9e815bb781ba29ccd057bb52dab Mon Sep 17 00:00:00 2001 From: ComfyUI Wiki Date: Tue, 23 Sep 2025 23:22:35 +0800 Subject: [PATCH 009/114] update template to 0.1.86 (#9998) * update template to 0.1.84 * update template to 0.1.85 * Update template to 0.1.86 --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 79187efaa..2980bebdd 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,5 +1,5 @@ comfyui-frontend-package==1.26.13 -comfyui-workflow-templates==0.1.81 +comfyui-workflow-templates==0.1.86 comfyui-embedded-docs==0.2.6 torch torchsde From e8087907995497c6971ee64bd5fa02cb49c1eda6 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Tue, 23 Sep 2025 18:36:47 +0300 Subject: [PATCH 010/114] feat(api-nodes): add wan t2i, t2v, i2v nodes (#9996) --- comfy_api_nodes/nodes_wan.py | 602 +++++++++++++++++++++++++++++++++++ nodes.py | 1 + 2 files changed, 603 insertions(+) create mode 100644 comfy_api_nodes/nodes_wan.py diff --git a/comfy_api_nodes/nodes_wan.py b/comfy_api_nodes/nodes_wan.py new file mode 100644 index 000000000..db5bd41c1 --- /dev/null +++ b/comfy_api_nodes/nodes_wan.py @@ -0,0 +1,602 @@ +import re +from typing import Optional, Type, Union +from typing_extensions import override + +import torch +from pydantic import BaseModel, Field +from comfy_api.latest import ComfyExtension, Input, io as comfy_io +from comfy_api_nodes.apis.client import ( + ApiEndpoint, + HttpMethod, + SynchronousOperation, + PollingOperation, + EmptyRequest, + R, + T, +) +from comfy_api_nodes.util.validation_utils import get_number_of_images, validate_audio_duration + +from comfy_api_nodes.apinode_utils import ( + download_url_to_image_tensor, + download_url_to_video_output, + tensor_to_base64_string, + audio_to_base64_string, +) + +class Text2ImageInputField(BaseModel): + prompt: str = Field(...) + negative_prompt: Optional[str] = Field(None) + + +class Text2VideoInputField(BaseModel): + prompt: str = Field(...) + negative_prompt: Optional[str] = Field(None) + audio_url: Optional[str] = Field(None) + + +class Image2VideoInputField(BaseModel): + prompt: str = Field(...) + negative_prompt: Optional[str] = Field(None) + img_url: str = Field(...) + audio_url: Optional[str] = Field(None) + + +class Txt2ImageParametersField(BaseModel): + size: str = Field(...) + n: int = Field(1, description="Number of images to generate.") # we support only value=1 + seed: int = Field(..., ge=0, le=2147483647) + prompt_extend: bool = Field(True) + watermark: bool = Field(True) + + +class Text2VideoParametersField(BaseModel): + size: str = Field(...) + seed: int = Field(..., ge=0, le=2147483647) + duration: int = Field(5, ge=5, le=10) + prompt_extend: bool = Field(True) + watermark: bool = Field(True) + audio: bool = Field(False, description="Should be audio generated automatically") + + +class Image2VideoParametersField(BaseModel): + resolution: str = Field(...) + seed: int = Field(..., ge=0, le=2147483647) + duration: int = Field(5, ge=5, le=10) + prompt_extend: bool = Field(True) + watermark: bool = Field(True) + audio: bool = Field(False, description="Should be audio generated automatically") + + +class Text2ImageTaskCreationRequest(BaseModel): + model: str = Field(...) + input: Text2ImageInputField = Field(...) + parameters: Txt2ImageParametersField = Field(...) + + +class Text2VideoTaskCreationRequest(BaseModel): + model: str = Field(...) + input: Text2VideoInputField = Field(...) + parameters: Text2VideoParametersField = Field(...) + + +class Image2VideoTaskCreationRequest(BaseModel): + model: str = Field(...) + input: Image2VideoInputField = Field(...) + parameters: Image2VideoParametersField = Field(...) + + +class TaskCreationOutputField(BaseModel): + task_id: str = Field(...) + task_status: str = Field(...) + + +class TaskCreationResponse(BaseModel): + output: Optional[TaskCreationOutputField] = Field(None) + request_id: str = Field(...) + code: Optional[str] = Field(None, description="The error code of the failed request.") + message: Optional[str] = Field(None, description="Details of the failed request.") + + +class TaskResult(BaseModel): + url: Optional[str] = Field(None) + code: Optional[str] = Field(None) + message: Optional[str] = Field(None) + + +class ImageTaskStatusOutputField(TaskCreationOutputField): + task_id: str = Field(...) + task_status: str = Field(...) + results: Optional[list[TaskResult]] = Field(None) + + +class VideoTaskStatusOutputField(TaskCreationOutputField): + task_id: str = Field(...) + task_status: str = Field(...) + video_url: Optional[str] = Field(None) + code: Optional[str] = Field(None) + message: Optional[str] = Field(None) + + +class ImageTaskStatusResponse(BaseModel): + output: Optional[ImageTaskStatusOutputField] = Field(None) + request_id: str = Field(...) + + +class VideoTaskStatusResponse(BaseModel): + output: Optional[VideoTaskStatusOutputField] = Field(None) + request_id: str = Field(...) + + +RES_IN_PARENS = re.compile(r'\((\d+)\s*[x×]\s*(\d+)\)') + + +async def process_task( + auth_kwargs: dict[str, str], + url: str, + request_model: Type[T], + response_model: Type[R], + payload: Union[Text2ImageTaskCreationRequest, Text2VideoTaskCreationRequest, Image2VideoTaskCreationRequest], + node_id: str, + estimated_duration: int, + poll_interval: int, +) -> Type[R]: + initial_response = await SynchronousOperation( + endpoint=ApiEndpoint( + path=url, + method=HttpMethod.POST, + request_model=request_model, + response_model=TaskCreationResponse, + ), + request=payload, + auth_kwargs=auth_kwargs, + ).execute() + + if not initial_response.output: + raise Exception(f"Unknown error occurred: {initial_response.code} - {initial_response.message}") + + return await PollingOperation( + poll_endpoint=ApiEndpoint( + path=f"/proxy/wan/api/v1/tasks/{initial_response.output.task_id}", + method=HttpMethod.GET, + request_model=EmptyRequest, + response_model=response_model, + ), + completed_statuses=["SUCCEEDED"], + failed_statuses=["FAILED", "CANCELED", "UNKNOWN"], + status_extractor=lambda x: x.output.task_status, + estimated_duration=estimated_duration, + poll_interval=poll_interval, + node_id=node_id, + auth_kwargs=auth_kwargs, + ).execute() + + +class WanTextToImageApi(comfy_io.ComfyNode): + @classmethod + def define_schema(cls): + return comfy_io.Schema( + node_id="WanTextToImageApi", + display_name="Wan Text to Image", + category="api node/image/Wan", + description="Generates image based on text prompt.", + inputs=[ + comfy_io.Combo.Input( + "model", + options=["wan2.5-t2i-preview"], + default="wan2.5-t2i-preview", + tooltip="Model to use.", + ), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt used to describe the elements and visual features, supports English/Chinese.", + ), + comfy_io.String.Input( + "negative_prompt", + multiline=True, + default="", + tooltip="Negative text prompt to guide what to avoid.", + optional=True, + ), + comfy_io.Int.Input( + "width", + default=1024, + min=768, + max=1440, + step=32, + optional=True, + ), + comfy_io.Int.Input( + "height", + default=1024, + min=768, + max=1440, + step=32, + optional=True, + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=2147483647, + step=1, + display_mode=comfy_io.NumberDisplay.number, + control_after_generate=True, + tooltip="Seed to use for generation.", + optional=True, + ), + comfy_io.Boolean.Input( + "prompt_extend", + default=True, + tooltip="Whether to enhance the prompt with AI assistance.", + optional=True, + ), + comfy_io.Boolean.Input( + "watermark", + default=True, + tooltip="Whether to add an \"AI generated\" watermark to the result.", + optional=True, + ), + ], + outputs=[ + comfy_io.Image.Output(), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, + model: str, + prompt: str, + negative_prompt: str = "", + width: int = 1024, + height: int = 1024, + seed: int = 0, + prompt_extend: bool = True, + watermark: bool = True, + ): + payload = Text2ImageTaskCreationRequest( + model=model, + input=Text2ImageInputField(prompt=prompt, negative_prompt=negative_prompt), + parameters=Txt2ImageParametersField( + size=f"{width}*{height}", + seed=seed, + prompt_extend=prompt_extend, + watermark=watermark, + ), + ) + response = await process_task( + { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + "/proxy/wan/api/v1/services/aigc/text2image/image-synthesis", + request_model=Text2ImageTaskCreationRequest, + response_model=ImageTaskStatusResponse, + payload=payload, + node_id=cls.hidden.unique_id, + estimated_duration=9, + poll_interval=3, + ) + return comfy_io.NodeOutput(await download_url_to_image_tensor(str(response.output.results[0].url))) + + +class WanTextToVideoApi(comfy_io.ComfyNode): + @classmethod + def define_schema(cls): + return comfy_io.Schema( + node_id="WanTextToVideoApi", + display_name="Wan Text to Video", + category="api node/video/Wan", + description="Generates video based on text prompt.", + inputs=[ + comfy_io.Combo.Input( + "model", + options=["wan2.5-t2v-preview"], + default="wan2.5-t2v-preview", + tooltip="Model to use.", + ), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt used to describe the elements and visual features, supports English/Chinese.", + ), + comfy_io.String.Input( + "negative_prompt", + multiline=True, + default="", + tooltip="Negative text prompt to guide what to avoid.", + optional=True, + ), + comfy_io.Combo.Input( + "size", + options=[ + "480p: 1:1 (624x624)", + "480p: 16:9 (832x480)", + "480p: 9:16 (480x832)", + "720p: 1:1 (960x960)", + "720p: 16:9 (1280x720)", + "720p: 9:16 (720x1280)", + "720p: 4:3 (1088x832)", + "720p: 3:4 (832x1088)", + "1080p: 1:1 (1440x1440)", + "1080p: 16:9 (1920x1080)", + "1080p: 9:16 (1080x1920)", + "1080p: 4:3 (1632x1248)", + "1080p: 3:4 (1248x1632)", + ], + default="480p: 1:1 (624x624)", + optional=True, + ), + comfy_io.Int.Input( + "duration", + default=5, + min=5, + max=10, + step=5, + display_mode=comfy_io.NumberDisplay.number, + tooltip="Available durations: 5 and 10 seconds", + optional=True, + ), + comfy_io.Audio.Input( + "audio", + optional=True, + tooltip="Audio must contain a clear, loud voice, without extraneous noise, background music.", + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=2147483647, + step=1, + display_mode=comfy_io.NumberDisplay.number, + control_after_generate=True, + tooltip="Seed to use for generation.", + optional=True, + ), + comfy_io.Boolean.Input( + "generate_audio", + default=False, + optional=True, + tooltip="If there is no audio input, generate audio automatically.", + ), + comfy_io.Boolean.Input( + "prompt_extend", + default=True, + tooltip="Whether to enhance the prompt with AI assistance.", + optional=True, + ), + comfy_io.Boolean.Input( + "watermark", + default=True, + tooltip="Whether to add an \"AI generated\" watermark to the result.", + optional=True, + ), + ], + outputs=[ + comfy_io.Video.Output(), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, + model: str, + prompt: str, + negative_prompt: str = "", + size: str = "480p: 1:1 (624x624)", + duration: int = 5, + audio: Optional[Input.Audio] = None, + seed: int = 0, + generate_audio: bool = False, + prompt_extend: bool = True, + watermark: bool = True, + ): + width, height = RES_IN_PARENS.search(size).groups() + audio_url = None + if audio is not None: + validate_audio_duration(audio, 3.0, 29.0) + audio_url = "data:audio/mp3;base64," + audio_to_base64_string(audio, "mp3", "libmp3lame") + payload = Text2VideoTaskCreationRequest( + model=model, + input=Text2VideoInputField(prompt=prompt, negative_prompt=negative_prompt, audio_url=audio_url), + parameters=Text2VideoParametersField( + size=f"{width}*{height}", + duration=duration, + seed=seed, + audio=generate_audio, + prompt_extend=prompt_extend, + watermark=watermark, + ), + ) + response = await process_task( + { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + "/proxy/wan/api/v1/services/aigc/video-generation/video-synthesis", + request_model=Text2VideoTaskCreationRequest, + response_model=VideoTaskStatusResponse, + payload=payload, + node_id=cls.hidden.unique_id, + estimated_duration=120 * int(duration / 5), + poll_interval=6, + ) + return comfy_io.NodeOutput(await download_url_to_video_output(response.output.video_url)) + + +class WanImageToVideoApi(comfy_io.ComfyNode): + @classmethod + def define_schema(cls): + return comfy_io.Schema( + node_id="WanImageToVideoApi", + display_name="Wan Image to Video", + category="api node/video/Wan", + description="Generates video based on the first frame and text prompt.", + inputs=[ + comfy_io.Combo.Input( + "model", + options=["wan2.5-i2v-preview"], + default="wan2.5-i2v-preview", + tooltip="Model to use.", + ), + comfy_io.Image.Input( + "image", + ), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt used to describe the elements and visual features, supports English/Chinese.", + ), + comfy_io.String.Input( + "negative_prompt", + multiline=True, + default="", + tooltip="Negative text prompt to guide what to avoid.", + optional=True, + ), + comfy_io.Combo.Input( + "resolution", + options=[ + "480P", + "720P", + "1080P", + ], + default="480P", + optional=True, + ), + comfy_io.Int.Input( + "duration", + default=5, + min=5, + max=10, + step=5, + display_mode=comfy_io.NumberDisplay.number, + tooltip="Available durations: 5 and 10 seconds", + optional=True, + ), + comfy_io.Audio.Input( + "audio", + optional=True, + tooltip="Audio must contain a clear, loud voice, without extraneous noise, background music.", + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=2147483647, + step=1, + display_mode=comfy_io.NumberDisplay.number, + control_after_generate=True, + tooltip="Seed to use for generation.", + optional=True, + ), + comfy_io.Boolean.Input( + "generate_audio", + default=False, + optional=True, + tooltip="If there is no audio input, generate audio automatically.", + ), + comfy_io.Boolean.Input( + "prompt_extend", + default=True, + tooltip="Whether to enhance the prompt with AI assistance.", + optional=True, + ), + comfy_io.Boolean.Input( + "watermark", + default=True, + tooltip="Whether to add an \"AI generated\" watermark to the result.", + optional=True, + ), + ], + outputs=[ + comfy_io.Video.Output(), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, + model: str, + image: torch.Tensor, + prompt: str, + negative_prompt: str = "", + resolution: str = "480P", + duration: int = 5, + audio: Optional[Input.Audio] = None, + seed: int = 0, + generate_audio: bool = False, + prompt_extend: bool = True, + watermark: bool = True, + ): + if get_number_of_images(image) != 1: + raise ValueError("Exactly one input image is required.") + image_url = "data:image/png;base64," + tensor_to_base64_string(image, total_pixels=2000*2000) + audio_url = None + if audio is not None: + validate_audio_duration(audio, 3.0, 29.0) + audio_url = "data:audio/mp3;base64," + audio_to_base64_string(audio, "mp3", "libmp3lame") + payload = Image2VideoTaskCreationRequest( + model=model, + input=Image2VideoInputField( + prompt=prompt, negative_prompt=negative_prompt, img_url=image_url, audio_url=audio_url + ), + parameters=Image2VideoParametersField( + resolution=resolution, + duration=duration, + seed=seed, + audio=generate_audio, + prompt_extend=prompt_extend, + watermark=watermark, + ), + ) + response = await process_task( + { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + "/proxy/wan/api/v1/services/aigc/video-generation/video-synthesis", + request_model=Image2VideoTaskCreationRequest, + response_model=VideoTaskStatusResponse, + payload=payload, + node_id=cls.hidden.unique_id, + estimated_duration=120 * int(duration / 5), + poll_interval=6, + ) + return comfy_io.NodeOutput(await download_url_to_video_output(response.output.video_url)) + + +class WanApiExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: + return [ + WanTextToImageApi, + WanTextToVideoApi, + WanImageToVideoApi, + ] + + +async def comfy_entrypoint() -> WanApiExtension: + return WanApiExtension() diff --git a/nodes.py b/nodes.py index 5a5fdcb8e..1a6784b68 100644 --- a/nodes.py +++ b/nodes.py @@ -2361,6 +2361,7 @@ async def init_builtin_api_nodes(): "nodes_rodin.py", "nodes_gemini.py", "nodes_vidu.py", + "nodes_wan.py", ] if not await load_custom_node(os.path.join(api_nodes_dir, "canary.py"), module_parent="comfy_api_nodes"): From b8730510db30c8858e1e5d8e126ef19eac395560 Mon Sep 17 00:00:00 2001 From: comfyanonymous Date: Tue, 23 Sep 2025 11:50:33 -0400 Subject: [PATCH 011/114] ComfyUI version 0.3.60 --- comfyui_version.py | 2 +- pyproject.toml | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/comfyui_version.py b/comfyui_version.py index ee58205f5..d469a8194 100644 --- a/comfyui_version.py +++ b/comfyui_version.py @@ -1,3 +1,3 @@ # This file is automatically generated by the build process when version is # updated in pyproject.toml. -__version__ = "0.3.59" +__version__ = "0.3.60" diff --git a/pyproject.toml b/pyproject.toml index a7fc1a5a6..7340c320b 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -1,6 +1,6 @@ [project] name = "ComfyUI" -version = "0.3.59" +version = "0.3.60" readme = "README.md" license = { file = "LICENSE" } requires-python = ">=3.9" From 341b4adefd308cbcf82c07effc255f2770b3b3e2 Mon Sep 17 00:00:00 2001 From: Changrz <51637999+WhiteGiven@users.noreply.github.com> Date: Thu, 25 Sep 2025 02:05:37 +0800 Subject: [PATCH 012/114] Rodin3D - add [Rodin3D Gen-2 generate] api-node (#9994) * update Rodin api node * update rodin3d gen2 api node * fix images limited bug --- comfy_api_nodes/apis/rodin_api.py | 3 +- comfy_api_nodes/nodes_rodin.py | 140 ++++++++++++++++++++++++------ 2 files changed, 117 insertions(+), 26 deletions(-) diff --git a/comfy_api_nodes/apis/rodin_api.py b/comfy_api_nodes/apis/rodin_api.py index b0cf171fa..02cf42c29 100644 --- a/comfy_api_nodes/apis/rodin_api.py +++ b/comfy_api_nodes/apis/rodin_api.py @@ -9,8 +9,9 @@ class Rodin3DGenerateRequest(BaseModel): seed: int = Field(..., description="seed_") tier: str = Field(..., description="Tier of generation.") material: str = Field(..., description="The material type.") - quality: str = Field(..., description="The generation quality of the mesh.") + quality_override: int = Field(..., description="The poly count of the mesh.") mesh_mode: str = Field(..., description="It controls the type of faces of generated models.") + TAPose: Optional[bool] = Field(None, description="") class GenerateJobsData(BaseModel): uuids: List[str] = Field(..., description="str LIST") diff --git a/comfy_api_nodes/nodes_rodin.py b/comfy_api_nodes/nodes_rodin.py index c89d087e5..1af393eba 100644 --- a/comfy_api_nodes/nodes_rodin.py +++ b/comfy_api_nodes/nodes_rodin.py @@ -121,10 +121,10 @@ class Rodin3DAPI: else: return "Generating" - async def create_generate_task(self, images=None, seed=1, material="PBR", quality="medium", tier="Regular", mesh_mode="Quad", **kwargs): + async def create_generate_task(self, images=None, seed=1, material="PBR", quality_override=18000, tier="Regular", mesh_mode="Quad", TAPose = False, **kwargs): if images is None: raise Exception("Rodin 3D generate requires at least 1 image.") - if len(images) >= 5: + if len(images) > 5: raise Exception("Rodin 3D generate requires up to 5 image.") path = "/proxy/rodin/api/v2/rodin" @@ -139,8 +139,9 @@ class Rodin3DAPI: seed=seed, tier=tier, material=material, - quality=quality, - mesh_mode=mesh_mode + quality_override=quality_override, + mesh_mode=mesh_mode, + TAPose=TAPose, ), files=[ ( @@ -211,23 +212,36 @@ class Rodin3DAPI: return await operation.execute() def get_quality_mode(self, poly_count): - if poly_count == "200K-Triangle": + polycount = poly_count.split("-") + poly = polycount[1] + count = polycount[0] + if poly == "Triangle": mesh_mode = "Raw" - quality = "medium" + elif poly == "Quad": + mesh_mode = "Quad" else: mesh_mode = "Quad" - if poly_count == "4K-Quad": - quality = "extra-low" - elif poly_count == "8K-Quad": - quality = "low" - elif poly_count == "18K-Quad": - quality = "medium" - elif poly_count == "50K-Quad": - quality = "high" - else: - quality = "medium" - return mesh_mode, quality + if count == "4K": + quality_override = 4000 + elif count == "8K": + quality_override = 8000 + elif count == "18K": + quality_override = 18000 + elif count == "50K": + quality_override = 50000 + elif count == "2K": + quality_override = 2000 + elif count == "20K": + quality_override = 20000 + elif count == "150K": + quality_override = 150000 + elif count == "500K": + quality_override = 500000 + else: + quality_override = 18000 + + return mesh_mode, quality_override async def download_files(self, url_list): save_path = os.path.join(comfy_paths.get_output_directory(), "Rodin3D", datetime.datetime.now().strftime("%Y-%m-%d_%H-%M-%S")) @@ -300,9 +314,9 @@ class Rodin3D_Regular(Rodin3DAPI): m_images = [] for i in range(num_images): m_images.append(Images[i]) - mesh_mode, quality = self.get_quality_mode(Polygon_count) + mesh_mode, quality_override = self.get_quality_mode(Polygon_count) task_uuid, subscription_key = await self.create_generate_task(images=m_images, seed=Seed, material=Material_Type, - quality=quality, tier=tier, mesh_mode=mesh_mode, + quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, **kwargs) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) @@ -346,9 +360,9 @@ class Rodin3D_Detail(Rodin3DAPI): m_images = [] for i in range(num_images): m_images.append(Images[i]) - mesh_mode, quality = self.get_quality_mode(Polygon_count) + mesh_mode, quality_override = self.get_quality_mode(Polygon_count) task_uuid, subscription_key = await self.create_generate_task(images=m_images, seed=Seed, material=Material_Type, - quality=quality, tier=tier, mesh_mode=mesh_mode, + quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, **kwargs) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) @@ -392,9 +406,9 @@ class Rodin3D_Smooth(Rodin3DAPI): m_images = [] for i in range(num_images): m_images.append(Images[i]) - mesh_mode, quality = self.get_quality_mode(Polygon_count) + mesh_mode, quality_override = self.get_quality_mode(Polygon_count) task_uuid, subscription_key = await self.create_generate_task(images=m_images, seed=Seed, material=Material_Type, - quality=quality, tier=tier, mesh_mode=mesh_mode, + quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, **kwargs) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) @@ -446,10 +460,10 @@ class Rodin3D_Sketch(Rodin3DAPI): for i in range(num_images): m_images.append(Images[i]) material_type = "PBR" - quality = "medium" + quality_override = 18000 mesh_mode = "Quad" task_uuid, subscription_key = await self.create_generate_task( - images=m_images, seed=Seed, material=material_type, quality=quality, tier=tier, mesh_mode=mesh_mode, **kwargs + images=m_images, seed=Seed, material=material_type, quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, **kwargs ) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) @@ -457,6 +471,80 @@ class Rodin3D_Sketch(Rodin3DAPI): return (model,) +class Rodin3D_Gen2(Rodin3DAPI): + @classmethod + def INPUT_TYPES(s): + return { + "required": { + "Images": + ( + IO.IMAGE, + { + "forceInput":True, + } + ) + }, + "optional": { + "Seed": ( + IO.INT, + { + "default":0, + "min":0, + "max":65535, + "display":"number" + } + ), + "Material_Type": ( + IO.COMBO, + { + "options": ["PBR", "Shaded"], + "default": "PBR" + } + ), + "Polygon_count": ( + IO.COMBO, + { + "options": ["4K-Quad", "8K-Quad", "18K-Quad", "50K-Quad", "2K-Triangle", "20K-Triangle", "150K-Triangle", "500K-Triangle"], + "default": "500K-Triangle" + } + ), + "TAPose": ( + IO.BOOLEAN, + { + "default": False, + } + ) + }, + "hidden": { + "auth_token": "AUTH_TOKEN_COMFY_ORG", + "comfy_api_key": "API_KEY_COMFY_ORG", + }, + } + + async def api_call( + self, + Images, + Seed, + Material_Type, + Polygon_count, + TAPose, + **kwargs + ): + tier = "Gen-2" + num_images = Images.shape[0] + m_images = [] + for i in range(num_images): + m_images.append(Images[i]) + mesh_mode, quality_override = self.get_quality_mode(Polygon_count) + task_uuid, subscription_key = await self.create_generate_task(images=m_images, seed=Seed, material=Material_Type, + quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, TAPose=TAPose, + **kwargs) + await self.poll_for_task_status(subscription_key, **kwargs) + download_list = await self.get_rodin_download_list(task_uuid, **kwargs) + model = await self.download_files(download_list) + + return (model,) + # A dictionary that contains all nodes you want to export with their names # NOTE: names should be globally unique NODE_CLASS_MAPPINGS = { @@ -464,6 +552,7 @@ NODE_CLASS_MAPPINGS = { "Rodin3D_Detail": Rodin3D_Detail, "Rodin3D_Smooth": Rodin3D_Smooth, "Rodin3D_Sketch": Rodin3D_Sketch, + "Rodin3D_Gen2": Rodin3D_Gen2, } # A dictionary that contains the friendly/humanly readable titles for the nodes @@ -472,4 +561,5 @@ NODE_DISPLAY_NAME_MAPPINGS = { "Rodin3D_Detail": "Rodin 3D Generate - Detail Generate", "Rodin3D_Smooth": "Rodin 3D Generate - Smooth Generate", "Rodin3D_Sketch": "Rodin 3D Generate - Sketch Generate", + "Rodin3D_Gen2": "Rodin 3D Generate - Gen-2 Generate", } From fd79d32f38fd24adca5a6e8214f05050f287c9db Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jukka=20Sepp=C3=A4nen?= <40791699+kijai@users.noreply.github.com> Date: Thu, 25 Sep 2025 01:59:29 +0300 Subject: [PATCH 013/114] Add new audio nodes (#9908) * Add new audio nodes - TrimAudioDuration - SplitAudioChannels - AudioConcat - AudioMerge - AudioAdjustVolume * Update nodes_audio.py * Add EmptyAudio -node * Change duration to Float (allows sub seconds) --- comfy_extras/nodes_audio.py | 223 ++++++++++++++++++++++++++++++++++++ 1 file changed, 223 insertions(+) diff --git a/comfy_extras/nodes_audio.py b/comfy_extras/nodes_audio.py index 3b23f65d8..51c8b9dd9 100644 --- a/comfy_extras/nodes_audio.py +++ b/comfy_extras/nodes_audio.py @@ -11,6 +11,7 @@ import json import random import hashlib import node_helpers +import logging from comfy.cli_args import args from comfy.comfy_types import FileLocator @@ -364,6 +365,216 @@ class RecordAudio: return (audio, ) +class TrimAudioDuration: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "audio": ("AUDIO",), + "start_index": ("FLOAT", {"default": 0.0, "min": -0xffffffffffffffff, "max": 0xffffffffffffffff, "step": 0.01, "tooltip": "Start time in seconds, can be negative to count from the end (supports sub-seconds)."}), + "duration": ("FLOAT", {"default": 60.0, "min": 0.0, "step": 0.01, "tooltip": "Duration in seconds"}), + }, + } + + FUNCTION = "trim" + RETURN_TYPES = ("AUDIO",) + CATEGORY = "audio" + DESCRIPTION = "Trim audio tensor into chosen time range." + + def trim(self, audio, start_index, duration): + waveform = audio["waveform"] + sample_rate = audio["sample_rate"] + audio_length = waveform.shape[-1] + + if start_index < 0: + start_frame = audio_length + int(round(start_index * sample_rate)) + else: + start_frame = int(round(start_index * sample_rate)) + start_frame = max(0, min(start_frame, audio_length - 1)) + + end_frame = start_frame + int(round(duration * sample_rate)) + end_frame = max(0, min(end_frame, audio_length)) + + if start_frame >= end_frame: + raise ValueError("AudioTrim: Start time must be less than end time and be within the audio length.") + + return ({"waveform": waveform[..., start_frame:end_frame], "sample_rate": sample_rate},) + + +class SplitAudioChannels: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "audio": ("AUDIO",), + }} + + RETURN_TYPES = ("AUDIO", "AUDIO") + RETURN_NAMES = ("left", "right") + FUNCTION = "separate" + CATEGORY = "audio" + DESCRIPTION = "Separates the audio into left and right channels." + + def separate(self, audio): + waveform = audio["waveform"] + sample_rate = audio["sample_rate"] + + if waveform.shape[1] != 2: + raise ValueError("AudioSplit: Input audio has only one channel.") + + left_channel = waveform[..., 0:1, :] + right_channel = waveform[..., 1:2, :] + + return ({"waveform": left_channel, "sample_rate": sample_rate}, {"waveform": right_channel, "sample_rate": sample_rate}) + + +def match_audio_sample_rates(waveform_1, sample_rate_1, waveform_2, sample_rate_2): + if sample_rate_1 != sample_rate_2: + if sample_rate_1 > sample_rate_2: + waveform_2 = torchaudio.functional.resample(waveform_2, sample_rate_2, sample_rate_1) + output_sample_rate = sample_rate_1 + logging.info(f"Resampling audio2 from {sample_rate_2}Hz to {sample_rate_1}Hz for merging.") + else: + waveform_1 = torchaudio.functional.resample(waveform_1, sample_rate_1, sample_rate_2) + output_sample_rate = sample_rate_2 + logging.info(f"Resampling audio1 from {sample_rate_1}Hz to {sample_rate_2}Hz for merging.") + else: + output_sample_rate = sample_rate_1 + return waveform_1, waveform_2, output_sample_rate + + +class AudioConcat: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "audio1": ("AUDIO",), + "audio2": ("AUDIO",), + "direction": (['after', 'before'], {"default": 'after', "tooltip": "Whether to append audio2 after or before audio1."}), + }} + + RETURN_TYPES = ("AUDIO",) + FUNCTION = "concat" + CATEGORY = "audio" + DESCRIPTION = "Concatenates the audio1 to audio2 in the specified direction." + + def concat(self, audio1, audio2, direction): + waveform_1 = audio1["waveform"] + waveform_2 = audio2["waveform"] + sample_rate_1 = audio1["sample_rate"] + sample_rate_2 = audio2["sample_rate"] + + if waveform_1.shape[1] == 1: + waveform_1 = waveform_1.repeat(1, 2, 1) + logging.info("AudioConcat: Converted mono audio1 to stereo by duplicating the channel.") + if waveform_2.shape[1] == 1: + waveform_2 = waveform_2.repeat(1, 2, 1) + logging.info("AudioConcat: Converted mono audio2 to stereo by duplicating the channel.") + + waveform_1, waveform_2, output_sample_rate = match_audio_sample_rates(waveform_1, sample_rate_1, waveform_2, sample_rate_2) + + if direction == 'after': + concatenated_audio = torch.cat((waveform_1, waveform_2), dim=2) + elif direction == 'before': + concatenated_audio = torch.cat((waveform_2, waveform_1), dim=2) + + return ({"waveform": concatenated_audio, "sample_rate": output_sample_rate},) + + +class AudioMerge: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "audio1": ("AUDIO",), + "audio2": ("AUDIO",), + "merge_method": (["add", "mean", "subtract", "multiply"], {"tooltip": "The method used to combine the audio waveforms."}), + }, + } + + FUNCTION = "merge" + RETURN_TYPES = ("AUDIO",) + CATEGORY = "audio" + DESCRIPTION = "Combine two audio tracks by overlaying their waveforms." + + def merge(self, audio1, audio2, merge_method): + waveform_1 = audio1["waveform"] + waveform_2 = audio2["waveform"] + sample_rate_1 = audio1["sample_rate"] + sample_rate_2 = audio2["sample_rate"] + + waveform_1, waveform_2, output_sample_rate = match_audio_sample_rates(waveform_1, sample_rate_1, waveform_2, sample_rate_2) + + length_1 = waveform_1.shape[-1] + length_2 = waveform_2.shape[-1] + + if length_2 > length_1: + logging.info(f"AudioMerge: Trimming audio2 from {length_2} to {length_1} samples to match audio1 length.") + waveform_2 = waveform_2[..., :length_1] + elif length_2 < length_1: + logging.info(f"AudioMerge: Padding audio2 from {length_2} to {length_1} samples to match audio1 length.") + pad_shape = list(waveform_2.shape) + pad_shape[-1] = length_1 - length_2 + pad_tensor = torch.zeros(pad_shape, dtype=waveform_2.dtype, device=waveform_2.device) + waveform_2 = torch.cat((waveform_2, pad_tensor), dim=-1) + + if merge_method == "add": + waveform = waveform_1 + waveform_2 + elif merge_method == "subtract": + waveform = waveform_1 - waveform_2 + elif merge_method == "multiply": + waveform = waveform_1 * waveform_2 + elif merge_method == "mean": + waveform = (waveform_1 + waveform_2) / 2 + + max_val = waveform.abs().max() + if max_val > 1.0: + waveform = waveform / max_val + + return ({"waveform": waveform, "sample_rate": output_sample_rate},) + + +class AudioAdjustVolume: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "audio": ("AUDIO",), + "volume": ("INT", {"default": 1.0, "min": -100, "max": 100, "tooltip": "Volume adjustment in decibels (dB). 0 = no change, +6 = double, -6 = half, etc"}), + }} + + RETURN_TYPES = ("AUDIO",) + FUNCTION = "adjust_volume" + CATEGORY = "audio" + + def adjust_volume(self, audio, volume): + if volume == 0: + return (audio,) + waveform = audio["waveform"] + sample_rate = audio["sample_rate"] + + gain = 10 ** (volume / 20) + waveform = waveform * gain + + return ({"waveform": waveform, "sample_rate": sample_rate},) + + +class EmptyAudio: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "duration": ("FLOAT", {"default": 60.0, "min": 0.0, "max": 0xffffffffffffffff, "step": 0.01, "tooltip": "Duration of the empty audio clip in seconds"}), + "sample_rate": ("INT", {"default": 44100, "tooltip": "Sample rate of the empty audio clip."}), + "channels": ("INT", {"default": 2, "min": 1, "max": 2, "tooltip": "Number of audio channels (1 for mono, 2 for stereo)."}), + }} + + RETURN_TYPES = ("AUDIO",) + FUNCTION = "create_empty_audio" + CATEGORY = "audio" + + def create_empty_audio(self, duration, sample_rate, channels): + num_samples = int(round(duration * sample_rate)) + waveform = torch.zeros((1, channels, num_samples), dtype=torch.float32) + return ({"waveform": waveform, "sample_rate": sample_rate},) + + NODE_CLASS_MAPPINGS = { "EmptyLatentAudio": EmptyLatentAudio, "VAEEncodeAudio": VAEEncodeAudio, @@ -375,6 +586,12 @@ NODE_CLASS_MAPPINGS = { "PreviewAudio": PreviewAudio, "ConditioningStableAudio": ConditioningStableAudio, "RecordAudio": RecordAudio, + "TrimAudioDuration": TrimAudioDuration, + "SplitAudioChannels": SplitAudioChannels, + "AudioConcat": AudioConcat, + "AudioMerge": AudioMerge, + "AudioAdjustVolume": AudioAdjustVolume, + "EmptyAudio": EmptyAudio, } NODE_DISPLAY_NAME_MAPPINGS = { @@ -387,4 +604,10 @@ NODE_DISPLAY_NAME_MAPPINGS = { "SaveAudioMP3": "Save Audio (MP3)", "SaveAudioOpus": "Save Audio (Opus)", "RecordAudio": "Record Audio", + "TrimAudioDuration": "Trim Audio Duration", + "SplitAudioChannels": "Split Audio Channels", + "AudioConcat": "Audio Concat", + "AudioMerge": "Audio Merge", + "AudioAdjustVolume": "Audio Adjust Volume", + "EmptyAudio": "Empty Audio", } From fccab99ec0fcd13e80fa59bc73bccff31f9450ca Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Wed, 24 Sep 2025 17:09:42 -0700 Subject: [PATCH 014/114] Fix issue with .view() in HuMo. (#10014) --- comfy/ldm/wan/model.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/comfy/ldm/wan/model.py b/comfy/ldm/wan/model.py index 2dac5980c..54616e6eb 100644 --- a/comfy/ldm/wan/model.py +++ b/comfy/ldm/wan/model.py @@ -1355,7 +1355,7 @@ class WanT2VCrossAttentionGather(WanSelfAttention): x = optimized_attention(q, k, v, heads=self.num_heads, skip_reshape=True, skip_output_reshape=True, transformer_options=transformer_options) - x = x.transpose(1, 2).view(b, -1, n, d).flatten(2) + x = x.transpose(1, 2).reshape(b, -1, n * d) x = self.o(x) return x From c8d2117f02bcad6d8316ffd8273bdc27adf83b44 Mon Sep 17 00:00:00 2001 From: Guy Niv <43928922+guyniv@users.noreply.github.com> Date: Thu, 25 Sep 2025 05:35:12 +0300 Subject: [PATCH 015/114] Fix memory leak by properly detaching model finalizer (#9979) When unloading models in load_models_gpu(), the model finalizer was not being explicitly detached, leading to a memory leak. This caused linear memory consumption increase over time as models are repeatedly loaded and unloaded. This change prevents orphaned finalizer references from accumulating in memory during model switching operations. --- comfy/model_management.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/comfy/model_management.py b/comfy/model_management.py index d880f1970..c5b817b62 100644 --- a/comfy/model_management.py +++ b/comfy/model_management.py @@ -645,7 +645,9 @@ def load_models_gpu(models, memory_required=0, force_patch_weights=False, minimu if loaded_model.model.is_clone(current_loaded_models[i].model): to_unload = [i] + to_unload for i in to_unload: - current_loaded_models.pop(i).model.detach(unpatch_all=False) + model_to_unload = current_loaded_models.pop(i) + model_to_unload.model.detach(unpatch_all=False) + model_to_unload.model_finalizer.detach() total_memory_required = {} for loaded_model in models_to_load: From ce4cb2389c8ce63cf8735f200b8672a2c1be0950 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Thu, 25 Sep 2025 14:20:13 -0700 Subject: [PATCH 016/114] Make LatentCompositeMasked work with basic video latents. (#10023) --- comfy_extras/nodes_mask.py | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-) diff --git a/comfy_extras/nodes_mask.py b/comfy_extras/nodes_mask.py index 2b0f8dd5d..a5e405008 100644 --- a/comfy_extras/nodes_mask.py +++ b/comfy_extras/nodes_mask.py @@ -12,35 +12,38 @@ from nodes import MAX_RESOLUTION def composite(destination, source, x, y, mask = None, multiplier = 8, resize_source = False): source = source.to(destination.device) if resize_source: - source = torch.nn.functional.interpolate(source, size=(destination.shape[2], destination.shape[3]), mode="bilinear") + source = torch.nn.functional.interpolate(source, size=(destination.shape[-2], destination.shape[-1]), mode="bilinear") source = comfy.utils.repeat_to_batch_size(source, destination.shape[0]) - x = max(-source.shape[3] * multiplier, min(x, destination.shape[3] * multiplier)) - y = max(-source.shape[2] * multiplier, min(y, destination.shape[2] * multiplier)) + x = max(-source.shape[-1] * multiplier, min(x, destination.shape[-1] * multiplier)) + y = max(-source.shape[-2] * multiplier, min(y, destination.shape[-2] * multiplier)) left, top = (x // multiplier, y // multiplier) - right, bottom = (left + source.shape[3], top + source.shape[2],) + right, bottom = (left + source.shape[-1], top + source.shape[-2],) if mask is None: mask = torch.ones_like(source) else: mask = mask.to(destination.device, copy=True) - mask = torch.nn.functional.interpolate(mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])), size=(source.shape[2], source.shape[3]), mode="bilinear") + mask = torch.nn.functional.interpolate(mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])), size=(source.shape[-2], source.shape[-1]), mode="bilinear") mask = comfy.utils.repeat_to_batch_size(mask, source.shape[0]) # calculate the bounds of the source that will be overlapping the destination # this prevents the source trying to overwrite latent pixels that are out of bounds # of the destination - visible_width, visible_height = (destination.shape[3] - left + min(0, x), destination.shape[2] - top + min(0, y),) + visible_width, visible_height = (destination.shape[-1] - left + min(0, x), destination.shape[-2] - top + min(0, y),) mask = mask[:, :, :visible_height, :visible_width] + if mask.ndim < source.ndim: + mask = mask.unsqueeze(1) + inverse_mask = torch.ones_like(mask) - mask - source_portion = mask * source[:, :, :visible_height, :visible_width] - destination_portion = inverse_mask * destination[:, :, top:bottom, left:right] + source_portion = mask * source[..., :visible_height, :visible_width] + destination_portion = inverse_mask * destination[..., top:bottom, left:right] - destination[:, :, top:bottom, left:right] = source_portion + destination_portion + destination[..., top:bottom, left:right] = source_portion + destination_portion return destination class LatentCompositeMasked: From 2b7f9a8196304badb5fe58e5c734e4b182ad0fdf Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Fri, 26 Sep 2025 11:12:43 -0700 Subject: [PATCH 017/114] Fix the failing unit test. (#10037) --- .github/workflows/test-unit.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/test-unit.yml b/.github/workflows/test-unit.yml index 78c918031..00caf5b8a 100644 --- a/.github/workflows/test-unit.yml +++ b/.github/workflows/test-unit.yml @@ -10,7 +10,7 @@ jobs: test: strategy: matrix: - os: [ubuntu-latest, windows-latest, macos-latest] + os: [ubuntu-latest, windows-2022, macos-latest] runs-on: ${{ matrix.os }} continue-on-error: true steps: From c4a46e943c12c7f3f6ac72f8fb51caad514ec9b6 Mon Sep 17 00:00:00 2001 From: Yoland Yan <4950057+yoland68@users.noreply.github.com> Date: Fri, 26 Sep 2025 14:08:16 -0700 Subject: [PATCH 018/114] Add @kosinkadink as code owner (#10041) Updated CODEOWNERS to include @kosinkadink as a code owner. --- CODEOWNERS | 24 +----------------------- 1 file changed, 1 insertion(+), 23 deletions(-) diff --git a/CODEOWNERS b/CODEOWNERS index c8acd66d5..b7aca9b26 100644 --- a/CODEOWNERS +++ b/CODEOWNERS @@ -1,25 +1,3 @@ # Admins * @comfyanonymous - -# Note: Github teams syntax cannot be used here as the repo is not owned by Comfy-Org. -# Inlined the team members for now. - -# Maintainers -*.md @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @Kosinkadink @christian-byrne @guill -/tests/ @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @Kosinkadink @christian-byrne @guill -/tests-unit/ @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @Kosinkadink @christian-byrne @guill -/notebooks/ @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @Kosinkadink @christian-byrne @guill -/script_examples/ @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @Kosinkadink @christian-byrne @guill -/.github/ @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @Kosinkadink @christian-byrne @guill -/requirements.txt @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @Kosinkadink @christian-byrne @guill -/pyproject.toml @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @Kosinkadink @christian-byrne @guill - -# Python web server -/api_server/ @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @christian-byrne @guill -/app/ @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @christian-byrne @guill -/utils/ @yoland68 @robinjhuang @webfiltered @pythongosssss @ltdrdata @christian-byrne @guill - -# Node developers -/comfy_extras/ @yoland68 @robinjhuang @pythongosssss @ltdrdata @Kosinkadink @webfiltered @christian-byrne @guill -/comfy/comfy_types/ @yoland68 @robinjhuang @pythongosssss @ltdrdata @Kosinkadink @webfiltered @christian-byrne @guill -/comfy_api_nodes/ @yoland68 @robinjhuang @pythongosssss @ltdrdata @Kosinkadink @webfiltered @christian-byrne @guill +* @kosinkadink From 76eb1d72c3e5bef51d6ca8a26bf996972d3f6d1a Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 00:10:49 +0300 Subject: [PATCH 019/114] convert nodes_rebatch.py to V3 schema (#9945) --- comfy_extras/nodes_rebatch.py | 97 ++++++++++++++++++++--------------- 1 file changed, 56 insertions(+), 41 deletions(-) diff --git a/comfy_extras/nodes_rebatch.py b/comfy_extras/nodes_rebatch.py index e29cb9ed1..5f4e82aef 100644 --- a/comfy_extras/nodes_rebatch.py +++ b/comfy_extras/nodes_rebatch.py @@ -1,18 +1,25 @@ +from typing_extensions import override import torch -class LatentRebatch: +from comfy_api.latest import ComfyExtension, io + + +class LatentRebatch(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "latents": ("LATENT",), - "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096}), - }} - RETURN_TYPES = ("LATENT",) - INPUT_IS_LIST = True - OUTPUT_IS_LIST = (True, ) - - FUNCTION = "rebatch" - - CATEGORY = "latent/batch" + def define_schema(cls): + return io.Schema( + node_id="RebatchLatents", + display_name="Rebatch Latents", + category="latent/batch", + is_input_list=True, + inputs=[ + io.Latent.Input("latents"), + io.Int.Input("batch_size", default=1, min=1, max=4096), + ], + outputs=[ + io.Latent.Output(is_output_list=True), + ], + ) @staticmethod def get_batch(latents, list_ind, offset): @@ -53,7 +60,8 @@ class LatentRebatch: result = [torch.cat((b1, b2)) if torch.is_tensor(b1) else b1 + b2 for b1, b2 in zip(batch1, batch2)] return result - def rebatch(self, latents, batch_size): + @classmethod + def execute(cls, latents, batch_size): batch_size = batch_size[0] output_list = [] @@ -63,24 +71,24 @@ class LatentRebatch: for i in range(len(latents)): # fetch new entry of list #samples, masks, indices = self.get_batch(latents, i) - next_batch = self.get_batch(latents, i, processed) + next_batch = cls.get_batch(latents, i, processed) processed += len(next_batch[2]) # set to current if current is None if current_batch[0] is None: current_batch = next_batch # add previous to list if dimensions do not match elif next_batch[0].shape[-1] != current_batch[0].shape[-1] or next_batch[0].shape[-2] != current_batch[0].shape[-2]: - sliced, _ = self.slice_batch(current_batch, 1, batch_size) + sliced, _ = cls.slice_batch(current_batch, 1, batch_size) output_list.append({'samples': sliced[0][0], 'noise_mask': sliced[1][0], 'batch_index': sliced[2][0]}) current_batch = next_batch # cat if everything checks out else: - current_batch = self.cat_batch(current_batch, next_batch) + current_batch = cls.cat_batch(current_batch, next_batch) # add to list if dimensions gone above target batch size if current_batch[0].shape[0] > batch_size: num = current_batch[0].shape[0] // batch_size - sliced, remainder = self.slice_batch(current_batch, num, batch_size) + sliced, remainder = cls.slice_batch(current_batch, num, batch_size) for i in range(num): output_list.append({'samples': sliced[0][i], 'noise_mask': sliced[1][i], 'batch_index': sliced[2][i]}) @@ -89,7 +97,7 @@ class LatentRebatch: #add remainder if current_batch[0] is not None: - sliced, _ = self.slice_batch(current_batch, 1, batch_size) + sliced, _ = cls.slice_batch(current_batch, 1, batch_size) output_list.append({'samples': sliced[0][0], 'noise_mask': sliced[1][0], 'batch_index': sliced[2][0]}) #get rid of empty masks @@ -97,23 +105,27 @@ class LatentRebatch: if s['noise_mask'].mean() == 1.0: del s['noise_mask'] - return (output_list,) + return io.NodeOutput(output_list) -class ImageRebatch: +class ImageRebatch(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "images": ("IMAGE",), - "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096}), - }} - RETURN_TYPES = ("IMAGE",) - INPUT_IS_LIST = True - OUTPUT_IS_LIST = (True, ) + def define_schema(cls): + return io.Schema( + node_id="RebatchImages", + display_name="Rebatch Images", + category="image/batch", + is_input_list=True, + inputs=[ + io.Image.Input("images"), + io.Int.Input("batch_size", default=1, min=1, max=4096), + ], + outputs=[ + io.Image.Output(is_output_list=True), + ], + ) - FUNCTION = "rebatch" - - CATEGORY = "image/batch" - - def rebatch(self, images, batch_size): + @classmethod + def execute(cls, images, batch_size): batch_size = batch_size[0] output_list = [] @@ -125,14 +137,17 @@ class ImageRebatch: for i in range(0, len(all_images), batch_size): output_list.append(torch.cat(all_images[i:i+batch_size], dim=0)) - return (output_list,) + return io.NodeOutput(output_list) -NODE_CLASS_MAPPINGS = { - "RebatchLatents": LatentRebatch, - "RebatchImages": ImageRebatch, -} -NODE_DISPLAY_NAME_MAPPINGS = { - "RebatchLatents": "Rebatch Latents", - "RebatchImages": "Rebatch Images", -} +class RebatchExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + LatentRebatch, + ImageRebatch, + ] + + +async def comfy_entrypoint() -> RebatchExtension: + return RebatchExtension() From 7ea173c1873ec22df6edabc80a912a08ae2d521b Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 00:12:04 +0300 Subject: [PATCH 020/114] convert nodes_fresca.py to V3 schema (#9951) --- comfy_extras/nodes_fresca.py | 61 +++++++++++++++++++++--------------- 1 file changed, 36 insertions(+), 25 deletions(-) diff --git a/comfy_extras/nodes_fresca.py b/comfy_extras/nodes_fresca.py index 65c2d0d0e..f308eb0c1 100644 --- a/comfy_extras/nodes_fresca.py +++ b/comfy_extras/nodes_fresca.py @@ -1,6 +1,8 @@ # Code based on https://github.com/WikiChao/FreSca (MIT License) import torch import torch.fft as fft +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io def Fourier_filter(x, scale_low=1.0, scale_high=1.5, freq_cutoff=20): @@ -51,25 +53,31 @@ def Fourier_filter(x, scale_low=1.0, scale_high=1.5, freq_cutoff=20): return x_filtered -class FreSca: +class FreSca(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return { - "required": { - "model": ("MODEL",), - "scale_low": ("FLOAT", {"default": 1.0, "min": 0, "max": 10, "step": 0.01, - "tooltip": "Scaling factor for low-frequency components"}), - "scale_high": ("FLOAT", {"default": 1.25, "min": 0, "max": 10, "step": 0.01, - "tooltip": "Scaling factor for high-frequency components"}), - "freq_cutoff": ("INT", {"default": 20, "min": 1, "max": 10000, "step": 1, - "tooltip": "Number of frequency indices around center to consider as low-frequency"}), - } - } - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" - CATEGORY = "_for_testing" - DESCRIPTION = "Applies frequency-dependent scaling to the guidance" - def patch(self, model, scale_low, scale_high, freq_cutoff): + def define_schema(cls): + return io.Schema( + node_id="FreSca", + display_name="FreSca", + category="_for_testing", + description="Applies frequency-dependent scaling to the guidance", + inputs=[ + io.Model.Input("model"), + io.Float.Input("scale_low", default=1.0, min=0, max=10, step=0.01, + tooltip="Scaling factor for low-frequency components"), + io.Float.Input("scale_high", default=1.25, min=0, max=10, step=0.01, + tooltip="Scaling factor for high-frequency components"), + io.Int.Input("freq_cutoff", default=20, min=1, max=10000, step=1, + tooltip="Number of frequency indices around center to consider as low-frequency"), + ], + outputs=[ + io.Model.Output(), + ], + is_experimental=True, + ) + + @classmethod + def execute(cls, model, scale_low, scale_high, freq_cutoff): def custom_cfg_function(args): conds_out = args["conds_out"] if len(conds_out) <= 1 or None in args["conds"][:2]: @@ -91,13 +99,16 @@ class FreSca: m = model.clone() m.set_model_sampler_pre_cfg_function(custom_cfg_function) - return (m,) + return io.NodeOutput(m) -NODE_CLASS_MAPPINGS = { - "FreSca": FreSca, -} +class FreScaExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + FreSca, + ] -NODE_DISPLAY_NAME_MAPPINGS = { - "FreSca": "FreSca", -} + +async def comfy_entrypoint() -> FreScaExtension: + return FreScaExtension() From 80718908a9ac1045ece84285ca568511dcc9bc46 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 00:12:38 +0300 Subject: [PATCH 021/114] convert nodes_sdupscale.py to V3 schema (#9943) --- comfy_extras/nodes_sdupscale.py | 54 +++++++++++++++++++++------------ 1 file changed, 35 insertions(+), 19 deletions(-) diff --git a/comfy_extras/nodes_sdupscale.py b/comfy_extras/nodes_sdupscale.py index bba67e8dd..31b373370 100644 --- a/comfy_extras/nodes_sdupscale.py +++ b/comfy_extras/nodes_sdupscale.py @@ -1,23 +1,31 @@ +from typing_extensions import override + import torch import comfy.utils +from comfy_api.latest import ComfyExtension, io -class SD_4XUpscale_Conditioning: +class SD_4XUpscale_Conditioning(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "images": ("IMAGE",), - "positive": ("CONDITIONING",), - "negative": ("CONDITIONING",), - "scale_ratio": ("FLOAT", {"default": 4.0, "min": 0.0, "max": 10.0, "step": 0.01}), - "noise_augmentation": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 1.0, "step": 0.001}), - }} - RETURN_TYPES = ("CONDITIONING", "CONDITIONING", "LATENT") - RETURN_NAMES = ("positive", "negative", "latent") + def define_schema(cls): + return io.Schema( + node_id="SD_4XUpscale_Conditioning", + category="conditioning/upscale_diffusion", + inputs=[ + io.Image.Input("images"), + io.Conditioning.Input("positive"), + io.Conditioning.Input("negative"), + io.Float.Input("scale_ratio", default=4.0, min=0.0, max=10.0, step=0.01), + io.Float.Input("noise_augmentation", default=0.0, min=0.0, max=1.0, step=0.001), + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + io.Latent.Output(display_name="latent"), + ], + ) - FUNCTION = "encode" - - CATEGORY = "conditioning/upscale_diffusion" - - def encode(self, images, positive, negative, scale_ratio, noise_augmentation): + @classmethod + def execute(cls, images, positive, negative, scale_ratio, noise_augmentation): width = max(1, round(images.shape[-2] * scale_ratio)) height = max(1, round(images.shape[-3] * scale_ratio)) @@ -39,8 +47,16 @@ class SD_4XUpscale_Conditioning: out_cn.append(n) latent = torch.zeros([images.shape[0], 4, height // 4, width // 4]) - return (out_cp, out_cn, {"samples":latent}) + return io.NodeOutput(out_cp, out_cn, {"samples":latent}) -NODE_CLASS_MAPPINGS = { - "SD_4XUpscale_Conditioning": SD_4XUpscale_Conditioning, -} + +class SdUpscaleExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + SD_4XUpscale_Conditioning, + ] + + +async def comfy_entrypoint() -> SdUpscaleExtension: + return SdUpscaleExtension() From a061b06321b4e91d05c7c436b1e9b188360c5377 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 00:13:05 +0300 Subject: [PATCH 022/114] convert nodes_tcfg.py to V3 schema (#9942) --- comfy_extras/nodes_tcfg.py | 51 +++++++++++++++++++++----------------- 1 file changed, 28 insertions(+), 23 deletions(-) diff --git a/comfy_extras/nodes_tcfg.py b/comfy_extras/nodes_tcfg.py index 35b89a73f..1a6767770 100644 --- a/comfy_extras/nodes_tcfg.py +++ b/comfy_extras/nodes_tcfg.py @@ -1,8 +1,9 @@ # TCFG: Tangential Damping Classifier-free Guidance - (arXiv: https://arxiv.org/abs/2503.18137) +from typing_extensions import override import torch -from comfy.comfy_types import IO, ComfyNodeABC, InputTypeDict +from comfy_api.latest import ComfyExtension, io def score_tangential_damping(cond_score: torch.Tensor, uncond_score: torch.Tensor) -> torch.Tensor: @@ -26,23 +27,24 @@ def score_tangential_damping(cond_score: torch.Tensor, uncond_score: torch.Tenso return uncond_score_td.reshape_as(uncond_score).to(uncond_score.dtype) -class TCFG(ComfyNodeABC): +class TCFG(io.ComfyNode): @classmethod - def INPUT_TYPES(cls) -> InputTypeDict: - return { - "required": { - "model": (IO.MODEL, {}), - } - } + def define_schema(cls): + return io.Schema( + node_id="TCFG", + display_name="Tangential Damping CFG", + category="advanced/guidance", + description="TCFG – Tangential Damping CFG (2503.18137)\n\nRefine the uncond (negative) to align with the cond (positive) for improving quality.", + inputs=[ + io.Model.Input("model"), + ], + outputs=[ + io.Model.Output(display_name="patched_model"), + ], + ) - RETURN_TYPES = (IO.MODEL,) - RETURN_NAMES = ("patched_model",) - FUNCTION = "patch" - - CATEGORY = "advanced/guidance" - DESCRIPTION = "TCFG – Tangential Damping CFG (2503.18137)\n\nRefine the uncond (negative) to align with the cond (positive) for improving quality." - - def patch(self, model): + @classmethod + def execute(cls, model): m = model.clone() def tangential_damping_cfg(args): @@ -59,13 +61,16 @@ class TCFG(ComfyNodeABC): return [cond_pred, uncond_pred_td] + conds_out[2:] m.set_model_sampler_pre_cfg_function(tangential_damping_cfg) - return (m,) + return io.NodeOutput(m) -NODE_CLASS_MAPPINGS = { - "TCFG": TCFG, -} +class TcfgExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + TCFG, + ] -NODE_DISPLAY_NAME_MAPPINGS = { - "TCFG": "Tangential Damping CFG", -} + +async def comfy_entrypoint() -> TcfgExtension: + return TcfgExtension() From d20576e6a3527d0763ba8d7a72c70ee66829690a Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 00:13:52 +0300 Subject: [PATCH 023/114] convert nodes_sag.py to V3 schema (#9940) --- comfy_extras/nodes_sag.py | 50 +++++++++++++++++++++++++-------------- 1 file changed, 32 insertions(+), 18 deletions(-) diff --git a/comfy_extras/nodes_sag.py b/comfy_extras/nodes_sag.py index 1bd8d7364..0f47db30b 100644 --- a/comfy_extras/nodes_sag.py +++ b/comfy_extras/nodes_sag.py @@ -2,10 +2,13 @@ import torch from torch import einsum import torch.nn.functional as F import math +from typing_extensions import override from einops import rearrange, repeat from comfy.ldm.modules.attention import optimized_attention import comfy.samplers +from comfy_api.latest import ComfyExtension, io + # from comfy/ldm/modules/attention.py # but modified to return attention scores as well as output @@ -104,19 +107,26 @@ def gaussian_blur_2d(img, kernel_size, sigma): img = F.conv2d(img, kernel2d, groups=img.shape[-3]) return img -class SelfAttentionGuidance: +class SelfAttentionGuidance(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "model": ("MODEL",), - "scale": ("FLOAT", {"default": 0.5, "min": -2.0, "max": 5.0, "step": 0.01}), - "blur_sigma": ("FLOAT", {"default": 2.0, "min": 0.0, "max": 10.0, "step": 0.1}), - }} - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" + def define_schema(cls): + return io.Schema( + node_id="SelfAttentionGuidance", + display_name="Self-Attention Guidance", + category="_for_testing", + inputs=[ + io.Model.Input("model"), + io.Float.Input("scale", default=0.5, min=-2.0, max=5.0, step=0.01), + io.Float.Input("blur_sigma", default=2.0, min=0.0, max=10.0, step=0.1), + ], + outputs=[ + io.Model.Output(), + ], + is_experimental=True, + ) - CATEGORY = "_for_testing" - - def patch(self, model, scale, blur_sigma): + @classmethod + def execute(cls, model, scale, blur_sigma): m = model.clone() attn_scores = None @@ -170,12 +180,16 @@ class SelfAttentionGuidance: # unet.mid_block.attentions[0].transformer_blocks[0].attn1.patch m.set_model_attn1_replace(attn_and_record, "middle", 0, 0) - return (m, ) + return io.NodeOutput(m) -NODE_CLASS_MAPPINGS = { - "SelfAttentionGuidance": SelfAttentionGuidance, -} -NODE_DISPLAY_NAME_MAPPINGS = { - "SelfAttentionGuidance": "Self-Attention Guidance", -} +class SagExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + SelfAttentionGuidance, + ] + + +async def comfy_entrypoint() -> SagExtension: + return SagExtension() From 2103e393350d297ef77497a1b14a8199d4a1f1b4 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 00:14:42 +0300 Subject: [PATCH 024/114] convert nodes_post_processing to V3 schema (#9491) --- comfy_extras/nodes_post_processing.py | 249 ++++++++++++-------------- 1 file changed, 111 insertions(+), 138 deletions(-) diff --git a/comfy_extras/nodes_post_processing.py b/comfy_extras/nodes_post_processing.py index ed7a07152..34c388a5a 100644 --- a/comfy_extras/nodes_post_processing.py +++ b/comfy_extras/nodes_post_processing.py @@ -1,3 +1,4 @@ +from typing_extensions import override import numpy as np import torch import torch.nn.functional as F @@ -7,33 +8,27 @@ import math import comfy.utils import comfy.model_management import node_helpers +from comfy_api.latest import ComfyExtension, io -class Blend: - def __init__(self): - pass +class Blend(io.ComfyNode): + @classmethod + def define_schema(cls): + return io.Schema( + node_id="ImageBlend", + category="image/postprocessing", + inputs=[ + io.Image.Input("image1"), + io.Image.Input("image2"), + io.Float.Input("blend_factor", default=0.5, min=0.0, max=1.0, step=0.01), + io.Combo.Input("blend_mode", options=["normal", "multiply", "screen", "overlay", "soft_light", "difference"]), + ], + outputs=[ + io.Image.Output(), + ], + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image1": ("IMAGE",), - "image2": ("IMAGE",), - "blend_factor": ("FLOAT", { - "default": 0.5, - "min": 0.0, - "max": 1.0, - "step": 0.01 - }), - "blend_mode": (["normal", "multiply", "screen", "overlay", "soft_light", "difference"],), - }, - } - - RETURN_TYPES = ("IMAGE",) - FUNCTION = "blend_images" - - CATEGORY = "image/postprocessing" - - def blend_images(self, image1: torch.Tensor, image2: torch.Tensor, blend_factor: float, blend_mode: str): + def execute(cls, image1: torch.Tensor, image2: torch.Tensor, blend_factor: float, blend_mode: str) -> io.NodeOutput: image1, image2 = node_helpers.image_alpha_fix(image1, image2) image2 = image2.to(image1.device) if image1.shape != image2.shape: @@ -41,12 +36,13 @@ class Blend: image2 = comfy.utils.common_upscale(image2, image1.shape[2], image1.shape[1], upscale_method='bicubic', crop='center') image2 = image2.permute(0, 2, 3, 1) - blended_image = self.blend_mode(image1, image2, blend_mode) + blended_image = cls.blend_mode(image1, image2, blend_mode) blended_image = image1 * (1 - blend_factor) + blended_image * blend_factor blended_image = torch.clamp(blended_image, 0, 1) - return (blended_image,) + return io.NodeOutput(blended_image) - def blend_mode(self, img1, img2, mode): + @classmethod + def blend_mode(cls, img1, img2, mode): if mode == "normal": return img2 elif mode == "multiply": @@ -56,13 +52,13 @@ class Blend: elif mode == "overlay": return torch.where(img1 <= 0.5, 2 * img1 * img2, 1 - 2 * (1 - img1) * (1 - img2)) elif mode == "soft_light": - return torch.where(img2 <= 0.5, img1 - (1 - 2 * img2) * img1 * (1 - img1), img1 + (2 * img2 - 1) * (self.g(img1) - img1)) + return torch.where(img2 <= 0.5, img1 - (1 - 2 * img2) * img1 * (1 - img1), img1 + (2 * img2 - 1) * (cls.g(img1) - img1)) elif mode == "difference": return img1 - img2 - else: - raise ValueError(f"Unsupported blend mode: {mode}") + raise ValueError(f"Unsupported blend mode: {mode}") - def g(self, x): + @classmethod + def g(cls, x): return torch.where(x <= 0.25, ((16 * x - 12) * x + 4) * x, torch.sqrt(x)) def gaussian_kernel(kernel_size: int, sigma: float, device=None): @@ -71,38 +67,26 @@ def gaussian_kernel(kernel_size: int, sigma: float, device=None): g = torch.exp(-(d * d) / (2.0 * sigma * sigma)) return g / g.sum() -class Blur: - def __init__(self): - pass +class Blur(io.ComfyNode): + @classmethod + def define_schema(cls): + return io.Schema( + node_id="ImageBlur", + category="image/postprocessing", + inputs=[ + io.Image.Input("image"), + io.Int.Input("blur_radius", default=1, min=1, max=31, step=1), + io.Float.Input("sigma", default=1.0, min=0.1, max=10.0, step=0.1), + ], + outputs=[ + io.Image.Output(), + ], + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": ("IMAGE",), - "blur_radius": ("INT", { - "default": 1, - "min": 1, - "max": 31, - "step": 1 - }), - "sigma": ("FLOAT", { - "default": 1.0, - "min": 0.1, - "max": 10.0, - "step": 0.1 - }), - }, - } - - RETURN_TYPES = ("IMAGE",) - FUNCTION = "blur" - - CATEGORY = "image/postprocessing" - - def blur(self, image: torch.Tensor, blur_radius: int, sigma: float): + def execute(cls, image: torch.Tensor, blur_radius: int, sigma: float) -> io.NodeOutput: if blur_radius == 0: - return (image,) + return io.NodeOutput(image) image = image.to(comfy.model_management.get_torch_device()) batch_size, height, width, channels = image.shape @@ -115,31 +99,24 @@ class Blur: blurred = F.conv2d(padded_image, kernel, padding=kernel_size // 2, groups=channels)[:,:,blur_radius:-blur_radius, blur_radius:-blur_radius] blurred = blurred.permute(0, 2, 3, 1) - return (blurred.to(comfy.model_management.intermediate_device()),) + return io.NodeOutput(blurred.to(comfy.model_management.intermediate_device())) -class Quantize: - def __init__(self): - pass +class Quantize(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": ("IMAGE",), - "colors": ("INT", { - "default": 256, - "min": 1, - "max": 256, - "step": 1 - }), - "dither": (["none", "floyd-steinberg", "bayer-2", "bayer-4", "bayer-8", "bayer-16"],), - }, - } - - RETURN_TYPES = ("IMAGE",) - FUNCTION = "quantize" - - CATEGORY = "image/postprocessing" + def define_schema(cls): + return io.Schema( + node_id="ImageQuantize", + category="image/postprocessing", + inputs=[ + io.Image.Input("image"), + io.Int.Input("colors", default=256, min=1, max=256, step=1), + io.Combo.Input("dither", options=["none", "floyd-steinberg", "bayer-2", "bayer-4", "bayer-8", "bayer-16"]), + ], + outputs=[ + io.Image.Output(), + ], + ) @staticmethod def bayer(im, pal_im, order): @@ -167,7 +144,8 @@ class Quantize: im = im.quantize(palette=pal_im, dither=Image.Dither.NONE) return im - def quantize(self, image: torch.Tensor, colors: int, dither: str): + @classmethod + def execute(cls, image: torch.Tensor, colors: int, dither: str) -> io.NodeOutput: batch_size, height, width, _ = image.shape result = torch.zeros_like(image) @@ -187,46 +165,29 @@ class Quantize: quantized_array = torch.tensor(np.array(quantized_image.convert("RGB"))).float() / 255 result[b] = quantized_array - return (result,) + return io.NodeOutput(result) -class Sharpen: - def __init__(self): - pass +class Sharpen(io.ComfyNode): + @classmethod + def define_schema(cls): + return io.Schema( + node_id="ImageSharpen", + category="image/postprocessing", + inputs=[ + io.Image.Input("image"), + io.Int.Input("sharpen_radius", default=1, min=1, max=31, step=1), + io.Float.Input("sigma", default=1.0, min=0.1, max=10.0, step=0.01), + io.Float.Input("alpha", default=1.0, min=0.0, max=5.0, step=0.01), + ], + outputs=[ + io.Image.Output(), + ], + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": ("IMAGE",), - "sharpen_radius": ("INT", { - "default": 1, - "min": 1, - "max": 31, - "step": 1 - }), - "sigma": ("FLOAT", { - "default": 1.0, - "min": 0.1, - "max": 10.0, - "step": 0.01 - }), - "alpha": ("FLOAT", { - "default": 1.0, - "min": 0.0, - "max": 5.0, - "step": 0.01 - }), - }, - } - - RETURN_TYPES = ("IMAGE",) - FUNCTION = "sharpen" - - CATEGORY = "image/postprocessing" - - def sharpen(self, image: torch.Tensor, sharpen_radius: int, sigma:float, alpha: float): + def execute(cls, image: torch.Tensor, sharpen_radius: int, sigma:float, alpha: float) -> io.NodeOutput: if sharpen_radius == 0: - return (image,) + return io.NodeOutput(image) batch_size, height, width, channels = image.shape image = image.to(comfy.model_management.get_torch_device()) @@ -245,23 +206,29 @@ class Sharpen: result = torch.clamp(sharpened, 0, 1) - return (result.to(comfy.model_management.intermediate_device()),) + return io.NodeOutput(result.to(comfy.model_management.intermediate_device())) -class ImageScaleToTotalPixels: +class ImageScaleToTotalPixels(io.ComfyNode): upscale_methods = ["nearest-exact", "bilinear", "area", "bicubic", "lanczos"] crop_methods = ["disabled", "center"] @classmethod - def INPUT_TYPES(s): - return {"required": { "image": ("IMAGE",), "upscale_method": (s.upscale_methods,), - "megapixels": ("FLOAT", {"default": 1.0, "min": 0.01, "max": 16.0, "step": 0.01}), - }} - RETURN_TYPES = ("IMAGE",) - FUNCTION = "upscale" + def define_schema(cls): + return io.Schema( + node_id="ImageScaleToTotalPixels", + category="image/upscaling", + inputs=[ + io.Image.Input("image"), + io.Combo.Input("upscale_method", options=cls.upscale_methods), + io.Float.Input("megapixels", default=1.0, min=0.01, max=16.0, step=0.01), + ], + outputs=[ + io.Image.Output(), + ], + ) - CATEGORY = "image/upscaling" - - def upscale(self, image, upscale_method, megapixels): + @classmethod + def execute(cls, image, upscale_method, megapixels) -> io.NodeOutput: samples = image.movedim(-1,1) total = int(megapixels * 1024 * 1024) @@ -271,12 +238,18 @@ class ImageScaleToTotalPixels: s = comfy.utils.common_upscale(samples, width, height, upscale_method, "disabled") s = s.movedim(1,-1) - return (s,) + return io.NodeOutput(s) -NODE_CLASS_MAPPINGS = { - "ImageBlend": Blend, - "ImageBlur": Blur, - "ImageQuantize": Quantize, - "ImageSharpen": Sharpen, - "ImageScaleToTotalPixels": ImageScaleToTotalPixels, -} +class PostProcessingExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + Blend, + Blur, + Quantize, + Sharpen, + ImageScaleToTotalPixels, + ] + +async def comfy_entrypoint() -> PostProcessingExtension: + return PostProcessingExtension() From cd66d72b464fd9d344baa426b50a5f0e5e512f99 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 00:15:44 +0300 Subject: [PATCH 025/114] convert CLIPTextEncodeSDXL nodes to V3 schema (#9716) --- comfy_extras/nodes_clip_sdxl.py | 93 +++++++++++++++++++-------------- 1 file changed, 55 insertions(+), 38 deletions(-) diff --git a/comfy_extras/nodes_clip_sdxl.py b/comfy_extras/nodes_clip_sdxl.py index 14269caf3..520ff0e3c 100644 --- a/comfy_extras/nodes_clip_sdxl.py +++ b/comfy_extras/nodes_clip_sdxl.py @@ -1,43 +1,52 @@ -from nodes import MAX_RESOLUTION +from typing_extensions import override -class CLIPTextEncodeSDXLRefiner: +import nodes +from comfy_api.latest import ComfyExtension, io + + +class CLIPTextEncodeSDXLRefiner(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { - "ascore": ("FLOAT", {"default": 6.0, "min": 0.0, "max": 1000.0, "step": 0.01}), - "width": ("INT", {"default": 1024.0, "min": 0, "max": MAX_RESOLUTION}), - "height": ("INT", {"default": 1024.0, "min": 0, "max": MAX_RESOLUTION}), - "text": ("STRING", {"multiline": True, "dynamicPrompts": True}), "clip": ("CLIP", ), - }} - RETURN_TYPES = ("CONDITIONING",) - FUNCTION = "encode" + def define_schema(cls): + return io.Schema( + node_id="CLIPTextEncodeSDXLRefiner", + category="advanced/conditioning", + inputs=[ + io.Float.Input("ascore", default=6.0, min=0.0, max=1000.0, step=0.01), + io.Int.Input("width", default=1024, min=0, max=nodes.MAX_RESOLUTION), + io.Int.Input("height", default=1024, min=0, max=nodes.MAX_RESOLUTION), + io.String.Input("text", multiline=True, dynamic_prompts=True), + io.Clip.Input("clip"), + ], + outputs=[io.Conditioning.Output()], + ) - CATEGORY = "advanced/conditioning" - - def encode(self, clip, ascore, width, height, text): + @classmethod + def execute(cls, clip, ascore, width, height, text) -> io.NodeOutput: tokens = clip.tokenize(text) - return (clip.encode_from_tokens_scheduled(tokens, add_dict={"aesthetic_score": ascore, "width": width, "height": height}), ) + return io.NodeOutput(clip.encode_from_tokens_scheduled(tokens, add_dict={"aesthetic_score": ascore, "width": width, "height": height})) -class CLIPTextEncodeSDXL: +class CLIPTextEncodeSDXL(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { - "clip": ("CLIP", ), - "width": ("INT", {"default": 1024.0, "min": 0, "max": MAX_RESOLUTION}), - "height": ("INT", {"default": 1024.0, "min": 0, "max": MAX_RESOLUTION}), - "crop_w": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION}), - "crop_h": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION}), - "target_width": ("INT", {"default": 1024.0, "min": 0, "max": MAX_RESOLUTION}), - "target_height": ("INT", {"default": 1024.0, "min": 0, "max": MAX_RESOLUTION}), - "text_g": ("STRING", {"multiline": True, "dynamicPrompts": True}), - "text_l": ("STRING", {"multiline": True, "dynamicPrompts": True}), - }} - RETURN_TYPES = ("CONDITIONING",) - FUNCTION = "encode" + def define_schema(cls): + return io.Schema( + node_id="CLIPTextEncodeSDXL", + category="advanced/conditioning", + inputs=[ + io.Clip.Input("clip"), + io.Int.Input("width", default=1024, min=0, max=nodes.MAX_RESOLUTION), + io.Int.Input("height", default=1024, min=0, max=nodes.MAX_RESOLUTION), + io.Int.Input("crop_w", default=0, min=0, max=nodes.MAX_RESOLUTION), + io.Int.Input("crop_h", default=0, min=0, max=nodes.MAX_RESOLUTION), + io.Int.Input("target_width", default=1024, min=0, max=nodes.MAX_RESOLUTION), + io.Int.Input("target_height", default=1024, min=0, max=nodes.MAX_RESOLUTION), + io.String.Input("text_g", multiline=True, dynamic_prompts=True), + io.String.Input("text_l", multiline=True, dynamic_prompts=True), + ], + outputs=[io.Conditioning.Output()], + ) - CATEGORY = "advanced/conditioning" - - def encode(self, clip, width, height, crop_w, crop_h, target_width, target_height, text_g, text_l): + @classmethod + def execute(cls, clip, width, height, crop_w, crop_h, target_width, target_height, text_g, text_l) -> io.NodeOutput: tokens = clip.tokenize(text_g) tokens["l"] = clip.tokenize(text_l)["l"] if len(tokens["l"]) != len(tokens["g"]): @@ -46,9 +55,17 @@ class CLIPTextEncodeSDXL: tokens["l"] += empty["l"] while len(tokens["l"]) > len(tokens["g"]): tokens["g"] += empty["g"] - return (clip.encode_from_tokens_scheduled(tokens, add_dict={"width": width, "height": height, "crop_w": crop_w, "crop_h": crop_h, "target_width": target_width, "target_height": target_height}), ) + return io.NodeOutput(clip.encode_from_tokens_scheduled(tokens, add_dict={"width": width, "height": height, "crop_w": crop_w, "crop_h": crop_h, "target_width": target_width, "target_height": target_height})) -NODE_CLASS_MAPPINGS = { - "CLIPTextEncodeSDXLRefiner": CLIPTextEncodeSDXLRefiner, - "CLIPTextEncodeSDXL": CLIPTextEncodeSDXL, -} + +class ClipSdxlExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + CLIPTextEncodeSDXLRefiner, + CLIPTextEncodeSDXL, + ] + + +async def comfy_entrypoint() -> ClipSdxlExtension: + return ClipSdxlExtension() From 1e098d61327e1c02c1a47b2626514474aa8e3c7e Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Fri, 26 Sep 2025 15:34:17 -0700 Subject: [PATCH 026/114] Don't add template to qwen2.5vl when template is in prompt. (#10043) Make the hunyuan image refiner template_end 36. --- comfy/text_encoders/hunyuan_image.py | 8 ++++- comfy/text_encoders/qwen_image.py | 46 +++++++++++++++++----------- 2 files changed, 35 insertions(+), 19 deletions(-) diff --git a/comfy/text_encoders/hunyuan_image.py b/comfy/text_encoders/hunyuan_image.py index 699eddc33..ff04726e1 100644 --- a/comfy/text_encoders/hunyuan_image.py +++ b/comfy/text_encoders/hunyuan_image.py @@ -63,7 +63,13 @@ class HunyuanImageTEModel(QwenImageTEModel): self.byt5_small = None def encode_token_weights(self, token_weight_pairs): - cond, p, extra = super().encode_token_weights(token_weight_pairs) + tok_pairs = token_weight_pairs["qwen25_7b"][0] + template_end = -1 + if tok_pairs[0][0] == 27: + if len(tok_pairs) > 36: # refiner prompt uses a fixed 36 template_end + template_end = 36 + + cond, p, extra = super().encode_token_weights(token_weight_pairs, template_end=template_end) if self.byt5_small is not None and "byt5" in token_weight_pairs: out = self.byt5_small.encode_token_weights(token_weight_pairs["byt5"]) extra["conditioning_byt5small"] = out[0] diff --git a/comfy/text_encoders/qwen_image.py b/comfy/text_encoders/qwen_image.py index 6646b1003..40fa67937 100644 --- a/comfy/text_encoders/qwen_image.py +++ b/comfy/text_encoders/qwen_image.py @@ -18,13 +18,22 @@ class QwenImageTokenizer(sd1_clip.SD1Tokenizer): self.llama_template_images = "<|im_start|>system\nDescribe the key features of the input image (color, shape, size, texture, objects, background), then explain how the user's text instruction should alter or modify the image. Generate a new image that meets the user's requirements while maintaining consistency with the original input where appropriate.<|im_end|>\n<|im_start|>user\n<|vision_start|><|image_pad|><|vision_end|>{}<|im_end|>\n<|im_start|>assistant\n" def tokenize_with_weights(self, text, return_word_ids=False, llama_template=None, images=[], **kwargs): - if llama_template is None: - if len(images) > 0: - llama_text = self.llama_template_images.format(text) - else: - llama_text = self.llama_template.format(text) + skip_template = False + if text.startswith('<|im_start|>'): + skip_template = True + if text.startswith('<|start_header_id|>'): + skip_template = True + + if skip_template: + llama_text = text else: - llama_text = llama_template.format(text) + if llama_template is None: + if len(images) > 0: + llama_text = self.llama_template_images.format(text) + else: + llama_text = self.llama_template.format(text) + else: + llama_text = llama_template.format(text) tokens = super().tokenize_with_weights(llama_text, return_word_ids=return_word_ids, disable_weights=True, **kwargs) key_name = next(iter(tokens)) embed_count = 0 @@ -47,22 +56,23 @@ class QwenImageTEModel(sd1_clip.SD1ClipModel): def __init__(self, device="cpu", dtype=None, model_options={}): super().__init__(device=device, dtype=dtype, name="qwen25_7b", clip_model=Qwen25_7BVLIModel, model_options=model_options) - def encode_token_weights(self, token_weight_pairs): + def encode_token_weights(self, token_weight_pairs, template_end=-1): out, pooled, extra = super().encode_token_weights(token_weight_pairs) tok_pairs = token_weight_pairs["qwen25_7b"][0] count_im_start = 0 - for i, v in enumerate(tok_pairs): - elem = v[0] - if not torch.is_tensor(elem): - if isinstance(elem, numbers.Integral): - if elem == 151644 and count_im_start < 2: - template_end = i - count_im_start += 1 + if template_end == -1: + for i, v in enumerate(tok_pairs): + elem = v[0] + if not torch.is_tensor(elem): + if isinstance(elem, numbers.Integral): + if elem == 151644 and count_im_start < 2: + template_end = i + count_im_start += 1 - if out.shape[1] > (template_end + 3): - if tok_pairs[template_end + 1][0] == 872: - if tok_pairs[template_end + 2][0] == 198: - template_end += 3 + if out.shape[1] > (template_end + 3): + if tok_pairs[template_end + 1][0] == 872: + if tok_pairs[template_end + 2][0] == 198: + template_end += 3 out = out[:, template_end:] From 196954ab8c55bc4ac48113686a57ce250677c7b5 Mon Sep 17 00:00:00 2001 From: Jedrzej Kosinski Date: Fri, 26 Sep 2025 19:55:03 -0700 Subject: [PATCH 027/114] Add 'input_cond' and 'input_uncond' to the args dictionary passed into sampler_cfg_function (#10044) --- comfy/samplers.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/comfy/samplers.py b/comfy/samplers.py index b3202cec6..c59e296a1 100755 --- a/comfy/samplers.py +++ b/comfy/samplers.py @@ -360,7 +360,7 @@ def calc_cond_uncond_batch(model, cond, uncond, x_in, timestep, model_options): def cfg_function(model, cond_pred, uncond_pred, cond_scale, x, timestep, model_options={}, cond=None, uncond=None): if "sampler_cfg_function" in model_options: args = {"cond": x - cond_pred, "uncond": x - uncond_pred, "cond_scale": cond_scale, "timestep": timestep, "input": x, "sigma": timestep, - "cond_denoised": cond_pred, "uncond_denoised": uncond_pred, "model": model, "model_options": model_options} + "cond_denoised": cond_pred, "uncond_denoised": uncond_pred, "model": model, "model_options": model_options, "input_cond": cond, "input_uncond": uncond} cfg_result = x - model_options["sampler_cfg_function"](args) else: cfg_result = uncond_pred + (cond_pred - uncond_pred) * cond_scale @@ -390,7 +390,7 @@ def sampling_function(model, x, timestep, uncond, cond, cond_scale, model_option for fn in model_options.get("sampler_pre_cfg_function", []): args = {"conds":conds, "conds_out": out, "cond_scale": cond_scale, "timestep": timestep, "input": x, "sigma": timestep, "model": model, "model_options": model_options} - out = fn(args) + out = fn(args) return cfg_function(model, out[0], out[1], cond_scale, x, timestep, model_options=model_options, cond=cond, uncond=uncond_) From 0572029fee48741a8cf34a8e4d485898c5ab5dfd Mon Sep 17 00:00:00 2001 From: ComfyUI Wiki Date: Sat, 27 Sep 2025 12:18:16 +0800 Subject: [PATCH 028/114] Update template to 0.1.88 (#10046) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 2980bebdd..b3f81e8fa 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,5 +1,5 @@ comfyui-frontend-package==1.26.13 -comfyui-workflow-templates==0.1.86 +comfyui-workflow-templates==0.1.88 comfyui-embedded-docs==0.2.6 torch torchsde From 255572188f79e5c58fa997bf73529021129459a9 Mon Sep 17 00:00:00 2001 From: Christian Byrne Date: Fri, 26 Sep 2025 21:29:13 -0700 Subject: [PATCH 029/114] Add workflow templates version tracking to system_stats (#9089) Adds installed and required workflow templates version information to the /system_stats endpoint, allowing the frontend to detect and notify users when their templates package is outdated. - Add get_installed_templates_version() and get_required_templates_version() methods to FrontendManager - Include templates version info in system_stats response - Add comprehensive unit tests for the new functionality --- app/frontend_management.py | 33 +++++++++ server.py | 4 ++ tests-unit/app_test/frontend_manager_test.py | 71 ++++++++++++++++++++ 3 files changed, 108 insertions(+) diff --git a/app/frontend_management.py b/app/frontend_management.py index 0bee73685..cce0c117d 100644 --- a/app/frontend_management.py +++ b/app/frontend_management.py @@ -42,6 +42,7 @@ def get_installed_frontend_version(): frontend_version_str = version("comfyui-frontend-package") return frontend_version_str + def get_required_frontend_version(): """Get the required frontend version from requirements.txt.""" try: @@ -63,6 +64,7 @@ def get_required_frontend_version(): logging.error(f"Error reading requirements.txt: {e}") return None + def check_frontend_version(): """Check if the frontend version is up to date.""" @@ -203,6 +205,37 @@ class FrontendManager: """Get the required frontend package version.""" return get_required_frontend_version() + @classmethod + def get_installed_templates_version(cls) -> str: + """Get the currently installed workflow templates package version.""" + try: + templates_version_str = version("comfyui-workflow-templates") + return templates_version_str + except Exception: + return None + + @classmethod + def get_required_templates_version(cls) -> str: + """Get the required workflow templates version from requirements.txt.""" + try: + with open(requirements_path, "r", encoding="utf-8") as f: + for line in f: + line = line.strip() + if line.startswith("comfyui-workflow-templates=="): + version_str = line.split("==")[-1] + if not is_valid_version(version_str): + logging.error(f"Invalid templates version format in requirements.txt: {version_str}") + return None + return version_str + logging.error("comfyui-workflow-templates not found in requirements.txt") + return None + except FileNotFoundError: + logging.error("requirements.txt not found. Cannot determine required templates version.") + return None + except Exception as e: + logging.error(f"Error reading requirements.txt: {e}") + return None + @classmethod def default_frontend_path(cls) -> str: try: diff --git a/server.py b/server.py index 603677397..80e9d3fa7 100644 --- a/server.py +++ b/server.py @@ -550,6 +550,8 @@ class PromptServer(): vram_total, torch_vram_total = comfy.model_management.get_total_memory(device, torch_total_too=True) vram_free, torch_vram_free = comfy.model_management.get_free_memory(device, torch_free_too=True) required_frontend_version = FrontendManager.get_required_frontend_version() + installed_templates_version = FrontendManager.get_installed_templates_version() + required_templates_version = FrontendManager.get_required_templates_version() system_stats = { "system": { @@ -558,6 +560,8 @@ class PromptServer(): "ram_free": ram_free, "comfyui_version": __version__, "required_frontend_version": required_frontend_version, + "installed_templates_version": installed_templates_version, + "required_templates_version": required_templates_version, "python_version": sys.version, "pytorch_version": comfy.model_management.torch_version, "embedded_python": os.path.split(os.path.split(sys.executable)[0])[1] == "python_embeded", diff --git a/tests-unit/app_test/frontend_manager_test.py b/tests-unit/app_test/frontend_manager_test.py index ce43ac564..643f04e72 100644 --- a/tests-unit/app_test/frontend_manager_test.py +++ b/tests-unit/app_test/frontend_manager_test.py @@ -205,3 +205,74 @@ numpy""" # Assert assert version is None + + +def test_get_templates_version(): + # Arrange + expected_version = "0.1.41" + mock_requirements_content = """torch +torchsde +comfyui-frontend-package==1.25.0 +comfyui-workflow-templates==0.1.41 +other-package==1.0.0 +numpy""" + + # Act + with patch("builtins.open", mock_open(read_data=mock_requirements_content)): + version = FrontendManager.get_required_templates_version() + + # Assert + assert version == expected_version + + +def test_get_templates_version_not_found(): + # Arrange + mock_requirements_content = """torch +torchsde +comfyui-frontend-package==1.25.0 +other-package==1.0.0 +numpy""" + + # Act + with patch("builtins.open", mock_open(read_data=mock_requirements_content)): + version = FrontendManager.get_required_templates_version() + + # Assert + assert version is None + + +def test_get_templates_version_invalid_semver(): + # Arrange + mock_requirements_content = """torch +torchsde +comfyui-workflow-templates==1.0.0.beta +other-package==1.0.0 +numpy""" + + # Act + with patch("builtins.open", mock_open(read_data=mock_requirements_content)): + version = FrontendManager.get_required_templates_version() + + # Assert + assert version is None + + +def test_get_installed_templates_version(): + # Arrange + expected_version = "0.1.40" + + # Act + with patch("app.frontend_management.version", return_value=expected_version): + version = FrontendManager.get_installed_templates_version() + + # Assert + assert version == expected_version + + +def test_get_installed_templates_version_not_installed(): + # Act + with patch("app.frontend_management.version", side_effect=Exception("Package not found")): + version = FrontendManager.get_installed_templates_version() + + # Assert + assert version is None From a9cf1cd249773632949bec2262f921f64378127f Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 09:13:05 +0300 Subject: [PATCH 030/114] convert nodes_hidream.py to V3 schema (#9946) --- comfy_extras/nodes_hidream.py | 88 +++++++++++++++++++++-------------- 1 file changed, 53 insertions(+), 35 deletions(-) diff --git a/comfy_extras/nodes_hidream.py b/comfy_extras/nodes_hidream.py index dfb98597b..eee683ee1 100644 --- a/comfy_extras/nodes_hidream.py +++ b/comfy_extras/nodes_hidream.py @@ -1,55 +1,73 @@ +from typing_extensions import override + import folder_paths import comfy.sd import comfy.model_management +from comfy_api.latest import ComfyExtension, io -class QuadrupleCLIPLoader: +class QuadrupleCLIPLoader(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "clip_name1": (folder_paths.get_filename_list("text_encoders"), ), - "clip_name2": (folder_paths.get_filename_list("text_encoders"), ), - "clip_name3": (folder_paths.get_filename_list("text_encoders"), ), - "clip_name4": (folder_paths.get_filename_list("text_encoders"), ) - }} - RETURN_TYPES = ("CLIP",) - FUNCTION = "load_clip" + def define_schema(cls): + return io.Schema( + node_id="QuadrupleCLIPLoader", + category="advanced/loaders", + description="[Recipes]\n\nhidream: long clip-l, long clip-g, t5xxl, llama_8b_3.1_instruct", + inputs=[ + io.Combo.Input("clip_name1", options=folder_paths.get_filename_list("text_encoders")), + io.Combo.Input("clip_name2", options=folder_paths.get_filename_list("text_encoders")), + io.Combo.Input("clip_name3", options=folder_paths.get_filename_list("text_encoders")), + io.Combo.Input("clip_name4", options=folder_paths.get_filename_list("text_encoders")), + ], + outputs=[ + io.Clip.Output(), + ] + ) - CATEGORY = "advanced/loaders" - - DESCRIPTION = "[Recipes]\n\nhidream: long clip-l, long clip-g, t5xxl, llama_8b_3.1_instruct" - - def load_clip(self, clip_name1, clip_name2, clip_name3, clip_name4): + @classmethod + def execute(cls, clip_name1, clip_name2, clip_name3, clip_name4): clip_path1 = folder_paths.get_full_path_or_raise("text_encoders", clip_name1) clip_path2 = folder_paths.get_full_path_or_raise("text_encoders", clip_name2) clip_path3 = folder_paths.get_full_path_or_raise("text_encoders", clip_name3) clip_path4 = folder_paths.get_full_path_or_raise("text_encoders", clip_name4) clip = comfy.sd.load_clip(ckpt_paths=[clip_path1, clip_path2, clip_path3, clip_path4], embedding_directory=folder_paths.get_folder_paths("embeddings")) - return (clip,) + return io.NodeOutput(clip) -class CLIPTextEncodeHiDream: +class CLIPTextEncodeHiDream(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { - "clip": ("CLIP", ), - "clip_l": ("STRING", {"multiline": True, "dynamicPrompts": True}), - "clip_g": ("STRING", {"multiline": True, "dynamicPrompts": True}), - "t5xxl": ("STRING", {"multiline": True, "dynamicPrompts": True}), - "llama": ("STRING", {"multiline": True, "dynamicPrompts": True}) - }} - RETURN_TYPES = ("CONDITIONING",) - FUNCTION = "encode" - - CATEGORY = "advanced/conditioning" - - def encode(self, clip, clip_l, clip_g, t5xxl, llama): + def define_schema(cls): + return io.Schema( + node_id="CLIPTextEncodeHiDream", + category="advanced/conditioning", + inputs=[ + io.Clip.Input("clip"), + io.String.Input("clip_l", multiline=True, dynamic_prompts=True), + io.String.Input("clip_g", multiline=True, dynamic_prompts=True), + io.String.Input("t5xxl", multiline=True, dynamic_prompts=True), + io.String.Input("llama", multiline=True, dynamic_prompts=True), + ], + outputs=[ + io.Conditioning.Output(), + ] + ) + @classmethod + def execute(cls, clip, clip_l, clip_g, t5xxl, llama): tokens = clip.tokenize(clip_g) tokens["l"] = clip.tokenize(clip_l)["l"] tokens["t5xxl"] = clip.tokenize(t5xxl)["t5xxl"] tokens["llama"] = clip.tokenize(llama)["llama"] - return (clip.encode_from_tokens_scheduled(tokens), ) + return io.NodeOutput(clip.encode_from_tokens_scheduled(tokens)) -NODE_CLASS_MAPPINGS = { - "QuadrupleCLIPLoader": QuadrupleCLIPLoader, - "CLIPTextEncodeHiDream": CLIPTextEncodeHiDream, -} + +class HiDreamExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + QuadrupleCLIPLoader, + CLIPTextEncodeHiDream, + ] + + +async def comfy_entrypoint() -> HiDreamExtension: + return HiDreamExtension() From 6b4b671ce7b6c412c2db9f9f83ff8e27dbcfd959 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 12:27:01 +0300 Subject: [PATCH 031/114] convert nodes_bfl.py to V3 schema (#10033) --- comfy_api_nodes/nodes_bfl.py | 1056 ++++++++++++++++------------------ 1 file changed, 489 insertions(+), 567 deletions(-) diff --git a/comfy_api_nodes/nodes_bfl.py b/comfy_api_nodes/nodes_bfl.py index c09be8d5b..77914021d 100644 --- a/comfy_api_nodes/nodes_bfl.py +++ b/comfy_api_nodes/nodes_bfl.py @@ -2,7 +2,8 @@ import asyncio import io from inspect import cleandoc from typing import Union, Optional -from comfy.comfy_types.node_typing import IO, ComfyNodeABC +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io as comfy_io from comfy_api_nodes.apis.bfl_api import ( BFLStatus, BFLFluxExpandImageRequest, @@ -130,7 +131,7 @@ def convert_image_to_base64(image: torch.Tensor): return base64.b64encode(img_byte_arr.getvalue()).decode() -class FluxProUltraImageNode(ComfyNodeABC): +class FluxProUltraImageNode(comfy_io.ComfyNode): """ Generates images using Flux Pro 1.1 Ultra via api based on prompt and resolution. """ @@ -141,71 +142,67 @@ class FluxProUltraImageNode(ComfyNodeABC): MAXIMUM_RATIO_STR = "4:1" @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation", - }, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="FluxProUltraImageNode", + display_name="Flux 1.1 [pro] Ultra Image", + category="api node/image/BFL", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation", ), - "prompt_upsampling": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", - }, + comfy_io.Boolean.Input( + "prompt_upsampling", + default=False, + tooltip="Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "The random seed used for creating the noise.", - }, + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="The random seed used for creating the noise.", ), - "aspect_ratio": ( - IO.STRING, - { - "default": "16:9", - "tooltip": "Aspect ratio of image; must be between 1:4 and 4:1.", - }, + comfy_io.String.Input( + "aspect_ratio", + default="16:9", + tooltip="Aspect ratio of image; must be between 1:4 and 4:1.", ), - "raw": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "When True, generate less processed, more natural-looking images.", - }, + comfy_io.Boolean.Input( + "raw", + default=False, + tooltip="When True, generate less processed, more natural-looking images.", ), - }, - "optional": { - "image_prompt": (IO.IMAGE,), - "image_prompt_strength": ( - IO.FLOAT, - { - "default": 0.1, - "min": 0.0, - "max": 1.0, - "step": 0.01, - "tooltip": "Blend between the prompt and the image prompt.", - }, + comfy_io.Image.Input( + "image_prompt", + optional=True, ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + comfy_io.Float.Input( + "image_prompt_strength", + default=0.1, + min=0.0, + max=1.0, + step=0.01, + tooltip="Blend between the prompt and the image prompt.", + optional=True, + ), + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def VALIDATE_INPUTS(cls, aspect_ratio: str): + def validate_inputs(cls, aspect_ratio: str): try: validate_aspect_ratio( aspect_ratio, @@ -218,14 +215,9 @@ class FluxProUltraImageNode(ComfyNodeABC): return str(e) return True - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/BFL" - - async def api_call( - self, + @classmethod + async def execute( + cls, prompt: str, aspect_ratio: str, prompt_upsampling=False, @@ -233,9 +225,7 @@ class FluxProUltraImageNode(ComfyNodeABC): seed=0, image_prompt=None, image_prompt_strength=0.1, - unique_id: Union[str, None] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: if image_prompt is None: validate_string(prompt, strip_whitespace=False) operation = SynchronousOperation( @@ -251,10 +241,10 @@ class FluxProUltraImageNode(ComfyNodeABC): seed=seed, aspect_ratio=validate_aspect_ratio( aspect_ratio, - minimum_ratio=self.MINIMUM_RATIO, - maximum_ratio=self.MAXIMUM_RATIO, - minimum_ratio_str=self.MINIMUM_RATIO_STR, - maximum_ratio_str=self.MAXIMUM_RATIO_STR, + minimum_ratio=cls.MINIMUM_RATIO, + maximum_ratio=cls.MAXIMUM_RATIO, + minimum_ratio_str=cls.MINIMUM_RATIO_STR, + maximum_ratio_str=cls.MAXIMUM_RATIO_STR, ), raw=raw, image_prompt=( @@ -266,13 +256,16 @@ class FluxProUltraImageNode(ComfyNodeABC): None if image_prompt is None else round(image_prompt_strength, 2) ), ), - auth_kwargs=kwargs, + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, ) - output_image = await handle_bfl_synchronous_operation(operation, node_id=unique_id) - return (output_image,) + output_image = await handle_bfl_synchronous_operation(operation, node_id=cls.hidden.unique_id) + return comfy_io.NodeOutput(output_image) -class FluxKontextProImageNode(ComfyNodeABC): +class FluxKontextProImageNode(comfy_io.ComfyNode): """ Edits images using Flux.1 Kontext [pro] via api based on prompt and aspect ratio. """ @@ -283,81 +276,73 @@ class FluxKontextProImageNode(ComfyNodeABC): MAXIMUM_RATIO_STR = "4:1" @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation - specify what and how to edit.", - }, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id=cls.NODE_ID, + display_name=cls.DISPLAY_NAME, + category="api node/image/BFL", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation - specify what and how to edit.", ), - "aspect_ratio": ( - IO.STRING, - { - "default": "16:9", - "tooltip": "Aspect ratio of image; must be between 1:4 and 4:1.", - }, + comfy_io.String.Input( + "aspect_ratio", + default="16:9", + tooltip="Aspect ratio of image; must be between 1:4 and 4:1.", ), - "guidance": ( - IO.FLOAT, - { - "default": 3.0, - "min": 0.1, - "max": 99.0, - "step": 0.1, - "tooltip": "Guidance strength for the image generation process" - }, + comfy_io.Float.Input( + "guidance", + default=3.0, + min=0.1, + max=99.0, + step=0.1, + tooltip="Guidance strength for the image generation process", ), - "steps": ( - IO.INT, - { - "default": 50, - "min": 1, - "max": 150, - "tooltip": "Number of steps for the image generation process" - }, + comfy_io.Int.Input( + "steps", + default=50, + min=1, + max=150, + tooltip="Number of steps for the image generation process", ), - "seed": ( - IO.INT, - { - "default": 1234, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "The random seed used for creating the noise.", - }, + comfy_io.Int.Input( + "seed", + default=1234, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="The random seed used for creating the noise.", ), - "prompt_upsampling": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", - }, + comfy_io.Boolean.Input( + "prompt_upsampling", + default=False, + tooltip="Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", ), - }, - "optional": { - "input_image": (IO.IMAGE,), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/BFL" + comfy_io.Image.Input( + "input_image", + optional=True, + ), + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) BFL_PATH = "/proxy/bfl/flux-kontext-pro/generate" + NODE_ID = "FluxKontextProImageNode" + DISPLAY_NAME = "Flux.1 Kontext [pro] Image" - async def api_call( - self, + @classmethod + async def execute( + cls, prompt: str, aspect_ratio: str, guidance: float, @@ -365,21 +350,19 @@ class FluxKontextProImageNode(ComfyNodeABC): input_image: Optional[torch.Tensor]=None, seed=0, prompt_upsampling=False, - unique_id: Union[str, None] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: aspect_ratio = validate_aspect_ratio( aspect_ratio, - minimum_ratio=self.MINIMUM_RATIO, - maximum_ratio=self.MAXIMUM_RATIO, - minimum_ratio_str=self.MINIMUM_RATIO_STR, - maximum_ratio_str=self.MAXIMUM_RATIO_STR, + minimum_ratio=cls.MINIMUM_RATIO, + maximum_ratio=cls.MAXIMUM_RATIO, + minimum_ratio_str=cls.MINIMUM_RATIO_STR, + maximum_ratio_str=cls.MAXIMUM_RATIO_STR, ) if input_image is None: validate_string(prompt, strip_whitespace=False) operation = SynchronousOperation( endpoint=ApiEndpoint( - path=self.BFL_PATH, + path=cls.BFL_PATH, method=HttpMethod.POST, request_model=BFLFluxKontextProGenerateRequest, response_model=BFLFluxProGenerateResponse, @@ -397,10 +380,13 @@ class FluxKontextProImageNode(ComfyNodeABC): else convert_image_to_base64(input_image) ) ), - auth_kwargs=kwargs, + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, ) - output_image = await handle_bfl_synchronous_operation(operation, node_id=unique_id) - return (output_image,) + output_image = await handle_bfl_synchronous_operation(operation, node_id=cls.hidden.unique_id) + return comfy_io.NodeOutput(output_image) class FluxKontextMaxImageNode(FluxKontextProImageNode): @@ -410,63 +396,60 @@ class FluxKontextMaxImageNode(FluxKontextProImageNode): DESCRIPTION = cleandoc(__doc__ or "") BFL_PATH = "/proxy/bfl/flux-kontext-max/generate" + NODE_ID = "FluxKontextMaxImageNode" + DISPLAY_NAME = "Flux.1 Kontext [max] Image" -class FluxProImageNode(ComfyNodeABC): +class FluxProImageNode(comfy_io.ComfyNode): """ Generates images synchronously based on prompt and resolution. """ @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation", - }, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="FluxProImageNode", + display_name="Flux 1.1 [pro] Image", + category="api node/image/BFL", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation", ), - "prompt_upsampling": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", - }, + comfy_io.Boolean.Input( + "prompt_upsampling", + default=False, + tooltip="Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", ), - "width": ( - IO.INT, - { - "default": 1024, - "min": 256, - "max": 1440, - "step": 32, - }, + comfy_io.Int.Input( + "width", + default=1024, + min=256, + max=1440, + step=32, ), - "height": ( - IO.INT, - { - "default": 768, - "min": 256, - "max": 1440, - "step": 32, - }, + comfy_io.Int.Input( + "height", + default=768, + min=256, + max=1440, + step=32, ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "The random seed used for creating the noise.", - }, + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="The random seed used for creating the noise.", + ), + comfy_io.Image.Input( + "image_prompt", + optional=True, ), - }, - "optional": { - "image_prompt": (IO.IMAGE,), # "image_prompt_strength": ( # IO.FLOAT, # { @@ -477,22 +460,19 @@ class FluxProImageNode(ComfyNodeABC): # "tooltip": "Blend between the prompt and the image prompt.", # }, # ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/BFL" - - async def api_call( - self, + @classmethod + async def execute( + cls, prompt: str, prompt_upsampling, width: int, @@ -500,9 +480,7 @@ class FluxProImageNode(ComfyNodeABC): seed=0, image_prompt=None, # image_prompt_strength=0.1, - unique_id: Union[str, None] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: image_prompt = ( image_prompt if image_prompt is None @@ -524,118 +502,103 @@ class FluxProImageNode(ComfyNodeABC): seed=seed, image_prompt=image_prompt, ), - auth_kwargs=kwargs, + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, ) - output_image = await handle_bfl_synchronous_operation(operation, node_id=unique_id) - return (output_image,) + output_image = await handle_bfl_synchronous_operation(operation, node_id=cls.hidden.unique_id) + return comfy_io.NodeOutput(output_image) -class FluxProExpandNode(ComfyNodeABC): +class FluxProExpandNode(comfy_io.ComfyNode): """ Outpaints image based on prompt. """ @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": (IO.IMAGE,), - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation", - }, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="FluxProExpandNode", + display_name="Flux.1 Expand Image", + category="api node/image/BFL", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("image"), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation", ), - "prompt_upsampling": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", - }, + comfy_io.Boolean.Input( + "prompt_upsampling", + default=False, + tooltip="Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", ), - "top": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 2048, - "tooltip": "Number of pixels to expand at the top of the image" - }, + comfy_io.Int.Input( + "top", + default=0, + min=0, + max=2048, + tooltip="Number of pixels to expand at the top of the image", ), - "bottom": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 2048, - "tooltip": "Number of pixels to expand at the bottom of the image" - }, + comfy_io.Int.Input( + "bottom", + default=0, + min=0, + max=2048, + tooltip="Number of pixels to expand at the bottom of the image", ), - "left": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 2048, - "tooltip": "Number of pixels to expand at the left side of the image" - }, + comfy_io.Int.Input( + "left", + default=0, + min=0, + max=2048, + tooltip="Number of pixels to expand at the left of the image", ), - "right": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 2048, - "tooltip": "Number of pixels to expand at the right side of the image" - }, + comfy_io.Int.Input( + "right", + default=0, + min=0, + max=2048, + tooltip="Number of pixels to expand at the right of the image", ), - "guidance": ( - IO.FLOAT, - { - "default": 60, - "min": 1.5, - "max": 100, - "tooltip": "Guidance strength for the image generation process" - }, + comfy_io.Float.Input( + "guidance", + default=60, + min=1.5, + max=100, + tooltip="Guidance strength for the image generation process", ), - "steps": ( - IO.INT, - { - "default": 50, - "min": 15, - "max": 50, - "tooltip": "Number of steps for the image generation process" - }, + comfy_io.Int.Input( + "steps", + default=50, + min=15, + max=50, + tooltip="Number of steps for the image generation process", ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "The random seed used for creating the noise.", - }, + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="The random seed used for creating the noise.", ), - }, - "optional": {}, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/BFL" - - async def api_call( - self, + @classmethod + async def execute( + cls, image: torch.Tensor, prompt: str, prompt_upsampling: bool, @@ -646,9 +609,7 @@ class FluxProExpandNode(ComfyNodeABC): steps: int, guidance: float, seed=0, - unique_id: Union[str, None] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: image = convert_image_to_base64(image) operation = SynchronousOperation( @@ -670,84 +631,77 @@ class FluxProExpandNode(ComfyNodeABC): seed=seed, image=image, ), - auth_kwargs=kwargs, + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, ) - output_image = await handle_bfl_synchronous_operation(operation, node_id=unique_id) - return (output_image,) + output_image = await handle_bfl_synchronous_operation(operation, node_id=cls.hidden.unique_id) + return comfy_io.NodeOutput(output_image) -class FluxProFillNode(ComfyNodeABC): +class FluxProFillNode(comfy_io.ComfyNode): """ Inpaints image based on mask and prompt. """ @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": (IO.IMAGE,), - "mask": (IO.MASK,), - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation", - }, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="FluxProFillNode", + display_name="Flux.1 Fill Image", + category="api node/image/BFL", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("image"), + comfy_io.Mask.Input("mask"), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation", ), - "prompt_upsampling": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", - }, + comfy_io.Boolean.Input( + "prompt_upsampling", + default=False, + tooltip="Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", ), - "guidance": ( - IO.FLOAT, - { - "default": 60, - "min": 1.5, - "max": 100, - "tooltip": "Guidance strength for the image generation process" - }, + comfy_io.Float.Input( + "guidance", + default=60, + min=1.5, + max=100, + tooltip="Guidance strength for the image generation process", ), - "steps": ( - IO.INT, - { - "default": 50, - "min": 15, - "max": 50, - "tooltip": "Number of steps for the image generation process" - }, + comfy_io.Int.Input( + "steps", + default=50, + min=15, + max=50, + tooltip="Number of steps for the image generation process", ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "The random seed used for creating the noise.", - }, + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="The random seed used for creating the noise.", ), - }, - "optional": {}, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/BFL" - - async def api_call( - self, + @classmethod + async def execute( + cls, image: torch.Tensor, mask: torch.Tensor, prompt: str, @@ -755,9 +709,7 @@ class FluxProFillNode(ComfyNodeABC): steps: int, guidance: float, seed=0, - unique_id: Union[str, None] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: # prepare mask mask = resize_mask_to_image(mask, image) mask = convert_image_to_base64(convert_mask_to_image(mask)) @@ -780,109 +732,96 @@ class FluxProFillNode(ComfyNodeABC): image=image, mask=mask, ), - auth_kwargs=kwargs, + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, ) - output_image = await handle_bfl_synchronous_operation(operation, node_id=unique_id) - return (output_image,) + output_image = await handle_bfl_synchronous_operation(operation, node_id=cls.hidden.unique_id) + return comfy_io.NodeOutput(output_image) -class FluxProCannyNode(ComfyNodeABC): +class FluxProCannyNode(comfy_io.ComfyNode): """ Generate image using a control image (canny). """ @classmethod - def INPUT_TYPES(s): - return { - "required": { - "control_image": (IO.IMAGE,), - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation", - }, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="FluxProCannyNode", + display_name="Flux.1 Canny Control Image", + category="api node/image/BFL", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("control_image"), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation", ), - "prompt_upsampling": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", - }, + comfy_io.Boolean.Input( + "prompt_upsampling", + default=False, + tooltip="Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", ), - "canny_low_threshold": ( - IO.FLOAT, - { - "default": 0.1, - "min": 0.01, - "max": 0.99, - "step": 0.01, - "tooltip": "Low threshold for Canny edge detection; ignored if skip_processing is True" - }, + comfy_io.Float.Input( + "canny_low_threshold", + default=0.1, + min=0.01, + max=0.99, + step=0.01, + tooltip="Low threshold for Canny edge detection; ignored if skip_processing is True", ), - "canny_high_threshold": ( - IO.FLOAT, - { - "default": 0.4, - "min": 0.01, - "max": 0.99, - "step": 0.01, - "tooltip": "High threshold for Canny edge detection; ignored if skip_processing is True" - }, + comfy_io.Float.Input( + "canny_high_threshold", + default=0.4, + min=0.01, + max=0.99, + step=0.01, + tooltip="High threshold for Canny edge detection; ignored if skip_processing is True", ), - "skip_preprocessing": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to skip preprocessing; set to True if control_image already is canny-fied, False if it is a raw image.", - }, + comfy_io.Boolean.Input( + "skip_preprocessing", + default=False, + tooltip="Whether to skip preprocessing; set to True if control_image already is canny-fied, False if it is a raw image.", ), - "guidance": ( - IO.FLOAT, - { - "default": 30, - "min": 1, - "max": 100, - "tooltip": "Guidance strength for the image generation process" - }, + comfy_io.Float.Input( + "guidance", + default=30, + min=1, + max=100, + tooltip="Guidance strength for the image generation process", ), - "steps": ( - IO.INT, - { - "default": 50, - "min": 15, - "max": 50, - "tooltip": "Number of steps for the image generation process" - }, + comfy_io.Int.Input( + "steps", + default=50, + min=15, + max=50, + tooltip="Number of steps for the image generation process", ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "The random seed used for creating the noise.", - }, + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="The random seed used for creating the noise.", ), - }, - "optional": {}, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/BFL" - - async def api_call( - self, + @classmethod + async def execute( + cls, control_image: torch.Tensor, prompt: str, prompt_upsampling: bool, @@ -892,9 +831,7 @@ class FluxProCannyNode(ComfyNodeABC): steps: int, guidance: float, seed=0, - unique_id: Union[str, None] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: control_image = convert_image_to_base64(control_image[:, :, :, :3]) preprocessed_image = None @@ -929,89 +866,80 @@ class FluxProCannyNode(ComfyNodeABC): canny_high_threshold=canny_high_threshold, preprocessed_image=preprocessed_image, ), - auth_kwargs=kwargs, + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, ) - output_image = await handle_bfl_synchronous_operation(operation, node_id=unique_id) - return (output_image,) + output_image = await handle_bfl_synchronous_operation(operation, node_id=cls.hidden.unique_id) + return comfy_io.NodeOutput(output_image) -class FluxProDepthNode(ComfyNodeABC): +class FluxProDepthNode(comfy_io.ComfyNode): """ Generate image using a control image (depth). """ @classmethod - def INPUT_TYPES(s): - return { - "required": { - "control_image": (IO.IMAGE,), - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation", - }, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="FluxProDepthNode", + display_name="Flux.1 Depth Control Image", + category="api node/image/BFL", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("control_image"), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation", ), - "prompt_upsampling": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", - }, + comfy_io.Boolean.Input( + "prompt_upsampling", + default=False, + tooltip="Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result).", ), - "skip_preprocessing": ( - IO.BOOLEAN, - { - "default": False, - "tooltip": "Whether to skip preprocessing; set to True if control_image already is depth-ified, False if it is a raw image.", - }, + comfy_io.Boolean.Input( + "skip_preprocessing", + default=False, + tooltip="Whether to skip preprocessing; set to True if control_image already is depth-ified, False if it is a raw image.", ), - "guidance": ( - IO.FLOAT, - { - "default": 15, - "min": 1, - "max": 100, - "tooltip": "Guidance strength for the image generation process" - }, + comfy_io.Float.Input( + "guidance", + default=15, + min=1, + max=100, + tooltip="Guidance strength for the image generation process", ), - "steps": ( - IO.INT, - { - "default": 50, - "min": 15, - "max": 50, - "tooltip": "Number of steps for the image generation process" - }, + comfy_io.Int.Input( + "steps", + default=50, + min=15, + max=50, + tooltip="Number of steps for the image generation process", ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "The random seed used for creating the noise.", - }, + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="The random seed used for creating the noise.", ), - }, - "optional": {}, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/BFL" - - async def api_call( - self, + @classmethod + async def execute( + cls, control_image: torch.Tensor, prompt: str, prompt_upsampling: bool, @@ -1019,9 +947,7 @@ class FluxProDepthNode(ComfyNodeABC): steps: int, guidance: float, seed=0, - unique_id: Union[str, None] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: control_image = convert_image_to_base64(control_image[:,:,:,:3]) preprocessed_image = None @@ -1045,33 +971,29 @@ class FluxProDepthNode(ComfyNodeABC): control_image=control_image, preprocessed_image=preprocessed_image, ), - auth_kwargs=kwargs, + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, ) - output_image = await handle_bfl_synchronous_operation(operation, node_id=unique_id) - return (output_image,) + output_image = await handle_bfl_synchronous_operation(operation, node_id=cls.hidden.unique_id) + return comfy_io.NodeOutput(output_image) -# A dictionary that contains all nodes you want to export with their names -# NOTE: names should be globally unique -NODE_CLASS_MAPPINGS = { - "FluxProUltraImageNode": FluxProUltraImageNode, - # "FluxProImageNode": FluxProImageNode, - "FluxKontextProImageNode": FluxKontextProImageNode, - "FluxKontextMaxImageNode": FluxKontextMaxImageNode, - "FluxProExpandNode": FluxProExpandNode, - "FluxProFillNode": FluxProFillNode, - "FluxProCannyNode": FluxProCannyNode, - "FluxProDepthNode": FluxProDepthNode, -} +class BFLExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: + return [ + FluxProUltraImageNode, + # FluxProImageNode, + FluxKontextProImageNode, + FluxKontextMaxImageNode, + FluxProExpandNode, + FluxProFillNode, + FluxProCannyNode, + FluxProDepthNode, + ] -# A dictionary that contains the friendly/humanly readable titles for the nodes -NODE_DISPLAY_NAME_MAPPINGS = { - "FluxProUltraImageNode": "Flux 1.1 [pro] Ultra Image", - # "FluxProImageNode": "Flux 1.1 [pro] Image", - "FluxKontextProImageNode": "Flux.1 Kontext [pro] Image", - "FluxKontextMaxImageNode": "Flux.1 Kontext [max] Image", - "FluxProExpandNode": "Flux.1 Expand Image", - "FluxProFillNode": "Flux.1 Fill Image", - "FluxProCannyNode": "Flux.1 Canny Control Image", - "FluxProDepthNode": "Flux.1 Depth Control Image", -} + +async def comfy_entrypoint() -> BFLExtension: + return BFLExtension() From bcfd80dd79ccfa77a7da69380795fbb55b65b1ba Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 12:28:11 +0300 Subject: [PATCH 032/114] convert nodes_luma.py to V3 schema (#10030) --- comfy_api_nodes/nodes_luma.py | 774 +++++++++++++++++----------------- 1 file changed, 396 insertions(+), 378 deletions(-) diff --git a/comfy_api_nodes/nodes_luma.py b/comfy_api_nodes/nodes_luma.py index b3c32bed5..9cd02ffd2 100644 --- a/comfy_api_nodes/nodes_luma.py +++ b/comfy_api_nodes/nodes_luma.py @@ -1,7 +1,8 @@ from __future__ import annotations from inspect import cleandoc from typing import Optional -from comfy.comfy_types.node_typing import IO, ComfyNodeABC +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io as comfy_io from comfy_api.input_impl.video_types import VideoFromFile from comfy_api_nodes.apis.luma_api import ( LumaImageModel, @@ -51,174 +52,186 @@ def image_result_url_extractor(response: LumaGeneration): def video_result_url_extractor(response: LumaGeneration): return response.assets.video if hasattr(response, "assets") and hasattr(response.assets, "video") else None -class LumaReferenceNode(ComfyNodeABC): +class LumaReferenceNode(comfy_io.ComfyNode): """ Holds an image and weight for use with Luma Generate Image node. """ - RETURN_TYPES = (LumaIO.LUMA_REF,) - RETURN_NAMES = ("luma_ref",) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "create_luma_reference" - CATEGORY = "api node/image/Luma" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="LumaReferenceNode", + display_name="Luma Reference", + category="api node/image/Luma", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input( + "image", + tooltip="Image to use as reference.", + ), + comfy_io.Float.Input( + "weight", + default=1.0, + min=0.0, + max=1.0, + step=0.01, + tooltip="Weight of image reference.", + ), + comfy_io.Custom(LumaIO.LUMA_REF).Input( + "luma_ref", + optional=True, + ), + ], + outputs=[comfy_io.Custom(LumaIO.LUMA_REF).Output(display_name="luma_ref")], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": ( - IO.IMAGE, - { - "tooltip": "Image to use as reference.", - }, - ), - "weight": ( - IO.FLOAT, - { - "default": 1.0, - "min": 0.0, - "max": 1.0, - "step": 0.01, - "tooltip": "Weight of image reference.", - }, - ), - }, - "optional": {"luma_ref": (LumaIO.LUMA_REF,)}, - } - - def create_luma_reference( - self, image: torch.Tensor, weight: float, luma_ref: LumaReferenceChain = None - ): + def execute( + cls, image: torch.Tensor, weight: float, luma_ref: LumaReferenceChain = None + ) -> comfy_io.NodeOutput: if luma_ref is not None: luma_ref = luma_ref.clone() else: luma_ref = LumaReferenceChain() luma_ref.add(LumaReference(image=image, weight=round(weight, 2))) - return (luma_ref,) + return comfy_io.NodeOutput(luma_ref) -class LumaConceptsNode(ComfyNodeABC): +class LumaConceptsNode(comfy_io.ComfyNode): """ Holds one or more Camera Concepts for use with Luma Text to Video and Luma Image to Video nodes. """ - RETURN_TYPES = (LumaIO.LUMA_CONCEPTS,) - RETURN_NAMES = ("luma_concepts",) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "create_concepts" - CATEGORY = "api node/video/Luma" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="LumaConceptsNode", + display_name="Luma Concepts", + category="api node/video/Luma", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Combo.Input( + "concept1", + options=get_luma_concepts(include_none=True), + ), + comfy_io.Combo.Input( + "concept2", + options=get_luma_concepts(include_none=True), + ), + comfy_io.Combo.Input( + "concept3", + options=get_luma_concepts(include_none=True), + ), + comfy_io.Combo.Input( + "concept4", + options=get_luma_concepts(include_none=True), + ), + comfy_io.Custom(LumaIO.LUMA_CONCEPTS).Input( + "luma_concepts", + tooltip="Optional Camera Concepts to add to the ones chosen here.", + optional=True, + ), + ], + outputs=[comfy_io.Custom(LumaIO.LUMA_CONCEPTS).Output(display_name="luma_concepts")], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "concept1": (get_luma_concepts(include_none=True),), - "concept2": (get_luma_concepts(include_none=True),), - "concept3": (get_luma_concepts(include_none=True),), - "concept4": (get_luma_concepts(include_none=True),), - }, - "optional": { - "luma_concepts": ( - LumaIO.LUMA_CONCEPTS, - { - "tooltip": "Optional Camera Concepts to add to the ones chosen here." - }, - ), - }, - } - - def create_concepts( - self, + def execute( + cls, concept1: str, concept2: str, concept3: str, concept4: str, luma_concepts: LumaConceptChain = None, - ): + ) -> comfy_io.NodeOutput: chain = LumaConceptChain(str_list=[concept1, concept2, concept3, concept4]) if luma_concepts is not None: chain = luma_concepts.clone_and_merge(chain) - return (chain,) + return comfy_io.NodeOutput(chain) -class LumaImageGenerationNode(ComfyNodeABC): +class LumaImageGenerationNode(comfy_io.ComfyNode): """ Generates images synchronously based on prompt and aspect ratio. """ - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/Luma" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="LumaImageNode", + display_name="Luma Text to Image", + category="api node/image/Luma", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation", + ), + comfy_io.Combo.Input( + "model", + options=[model.value for model in LumaImageModel], + ), + comfy_io.Combo.Input( + "aspect_ratio", + options=[ratio.value for ratio in LumaAspectRatio], + default=LumaAspectRatio.ratio_16_9, + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="Seed to determine if node should re-run; actual results are nondeterministic regardless of seed.", + ), + comfy_io.Float.Input( + "style_image_weight", + default=1.0, + min=0.0, + max=1.0, + step=0.01, + tooltip="Weight of style image. Ignored if no style_image provided.", + ), + comfy_io.Custom(LumaIO.LUMA_REF).Input( + "image_luma_ref", + tooltip="Luma Reference node connection to influence generation with input images; up to 4 images can be considered.", + optional=True, + ), + comfy_io.Image.Input( + "style_image", + tooltip="Style reference image; only 1 image will be used.", + optional=True, + ), + comfy_io.Image.Input( + "character_image", + tooltip="Character reference images; can be a batch of multiple, up to 4 images can be considered.", + optional=True, + ), + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation", - }, - ), - "model": ([model.value for model in LumaImageModel],), - "aspect_ratio": ( - [ratio.value for ratio in LumaAspectRatio], - { - "default": LumaAspectRatio.ratio_16_9, - }, - ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "Seed to determine if node should re-run; actual results are nondeterministic regardless of seed.", - }, - ), - "style_image_weight": ( - IO.FLOAT, - { - "default": 1.0, - "min": 0.0, - "max": 1.0, - "step": 0.01, - "tooltip": "Weight of style image. Ignored if no style_image provided.", - }, - ), - }, - "optional": { - "image_luma_ref": ( - LumaIO.LUMA_REF, - { - "tooltip": "Luma Reference node connection to influence generation with input images; up to 4 images can be considered." - }, - ), - "style_image": ( - IO.IMAGE, - {"tooltip": "Style reference image; only 1 image will be used."}, - ), - "character_image": ( - IO.IMAGE, - { - "tooltip": "Character reference images; can be a batch of multiple, up to 4 images can be considered." - }, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - async def api_call( - self, + async def execute( + cls, prompt: str, model: str, aspect_ratio: str, @@ -227,27 +240,29 @@ class LumaImageGenerationNode(ComfyNodeABC): image_luma_ref: LumaReferenceChain = None, style_image: torch.Tensor = None, character_image: torch.Tensor = None, - unique_id: str = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: validate_string(prompt, strip_whitespace=True, min_length=3) + auth_kwargs = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } # handle image_luma_ref api_image_ref = None if image_luma_ref is not None: - api_image_ref = await self._convert_luma_refs( - image_luma_ref, max_refs=4, auth_kwargs=kwargs, + api_image_ref = await cls._convert_luma_refs( + image_luma_ref, max_refs=4, auth_kwargs=auth_kwargs, ) # handle style_luma_ref api_style_ref = None if style_image is not None: - api_style_ref = await self._convert_style_image( - style_image, weight=style_image_weight, auth_kwargs=kwargs, + api_style_ref = await cls._convert_style_image( + style_image, weight=style_image_weight, auth_kwargs=auth_kwargs, ) # handle character_ref images character_ref = None if character_image is not None: download_urls = await upload_images_to_comfyapi( - character_image, max_images=4, auth_kwargs=kwargs, + character_image, max_images=4, auth_kwargs=auth_kwargs, ) character_ref = LumaCharacterRef( identity0=LumaImageIdentity(images=download_urls) @@ -268,7 +283,7 @@ class LumaImageGenerationNode(ComfyNodeABC): style_ref=api_style_ref, character_ref=character_ref, ), - auth_kwargs=kwargs, + auth_kwargs=auth_kwargs, ) response_api: LumaGeneration = await operation.execute() @@ -283,18 +298,19 @@ class LumaImageGenerationNode(ComfyNodeABC): failed_statuses=[LumaState.failed], status_extractor=lambda x: x.state, result_url_extractor=image_result_url_extractor, - node_id=unique_id, - auth_kwargs=kwargs, + node_id=cls.hidden.unique_id, + auth_kwargs=auth_kwargs, ) response_poll = await operation.execute() async with aiohttp.ClientSession() as session: async with session.get(response_poll.assets.image) as img_response: img = process_image_response(await img_response.content.read()) - return (img,) + return comfy_io.NodeOutput(img) + @classmethod async def _convert_luma_refs( - self, luma_ref: LumaReferenceChain, max_refs: int, auth_kwargs: Optional[dict[str,str]] = None + cls, luma_ref: LumaReferenceChain, max_refs: int, auth_kwargs: Optional[dict[str,str]] = None ): luma_urls = [] ref_count = 0 @@ -308,82 +324,84 @@ class LumaImageGenerationNode(ComfyNodeABC): break return luma_ref.create_api_model(download_urls=luma_urls, max_refs=max_refs) + @classmethod async def _convert_style_image( - self, style_image: torch.Tensor, weight: float, auth_kwargs: Optional[dict[str,str]] = None + cls, style_image: torch.Tensor, weight: float, auth_kwargs: Optional[dict[str,str]] = None ): chain = LumaReferenceChain( first_ref=LumaReference(image=style_image, weight=weight) ) - return await self._convert_luma_refs(chain, max_refs=1, auth_kwargs=auth_kwargs) + return await cls._convert_luma_refs(chain, max_refs=1, auth_kwargs=auth_kwargs) -class LumaImageModifyNode(ComfyNodeABC): +class LumaImageModifyNode(comfy_io.ComfyNode): """ Modifies images synchronously based on prompt and aspect ratio. """ - RETURN_TYPES = (IO.IMAGE,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/image/Luma" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="LumaImageModifyNode", + display_name="Luma Image to Image", + category="api node/image/Luma", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input( + "image", + ), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the image generation", + ), + comfy_io.Float.Input( + "image_weight", + default=0.1, + min=0.0, + max=0.98, + step=0.01, + tooltip="Weight of the image; the closer to 1.0, the less the image will be modified.", + ), + comfy_io.Combo.Input( + "model", + options=[model.value for model in LumaImageModel], + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="Seed to determine if node should re-run; actual results are nondeterministic regardless of seed.", + ), + ], + outputs=[comfy_io.Image.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": (IO.IMAGE,), - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the image generation", - }, - ), - "image_weight": ( - IO.FLOAT, - { - "default": 0.1, - "min": 0.0, - "max": 0.98, - "step": 0.01, - "tooltip": "Weight of the image; the closer to 1.0, the less the image will be modified.", - }, - ), - "model": ([model.value for model in LumaImageModel],), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "Seed to determine if node should re-run; actual results are nondeterministic regardless of seed.", - }, - ), - }, - "optional": {}, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - async def api_call( - self, + async def execute( + cls, prompt: str, model: str, image: torch.Tensor, image_weight: float, seed, - unique_id: str = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: + auth_kwargs = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } # first, upload image download_urls = await upload_images_to_comfyapi( - image, max_images=1, auth_kwargs=kwargs, + image, max_images=1, auth_kwargs=auth_kwargs, ) image_url = download_urls[0] # next, make Luma call with download url provided @@ -401,7 +419,7 @@ class LumaImageModifyNode(ComfyNodeABC): url=image_url, weight=round(max(min(1.0-image_weight, 0.98), 0.0), 2) ), ), - auth_kwargs=kwargs, + auth_kwargs=auth_kwargs, ) response_api: LumaGeneration = await operation.execute() @@ -416,88 +434,84 @@ class LumaImageModifyNode(ComfyNodeABC): failed_statuses=[LumaState.failed], status_extractor=lambda x: x.state, result_url_extractor=image_result_url_extractor, - node_id=unique_id, - auth_kwargs=kwargs, + node_id=cls.hidden.unique_id, + auth_kwargs=auth_kwargs, ) response_poll = await operation.execute() async with aiohttp.ClientSession() as session: async with session.get(response_poll.assets.image) as img_response: img = process_image_response(await img_response.content.read()) - return (img,) + return comfy_io.NodeOutput(img) -class LumaTextToVideoGenerationNode(ComfyNodeABC): +class LumaTextToVideoGenerationNode(comfy_io.ComfyNode): """ Generates videos synchronously based on prompt and output_size. """ - RETURN_TYPES = (IO.VIDEO,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/video/Luma" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="LumaVideoNode", + display_name="Luma Text to Video", + category="api node/video/Luma", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the video generation", + ), + comfy_io.Combo.Input( + "model", + options=[model.value for model in LumaVideoModel], + ), + comfy_io.Combo.Input( + "aspect_ratio", + options=[ratio.value for ratio in LumaAspectRatio], + default=LumaAspectRatio.ratio_16_9, + ), + comfy_io.Combo.Input( + "resolution", + options=[resolution.value for resolution in LumaVideoOutputResolution], + default=LumaVideoOutputResolution.res_540p, + ), + comfy_io.Combo.Input( + "duration", + options=[dur.value for dur in LumaVideoModelOutputDuration], + ), + comfy_io.Boolean.Input( + "loop", + default=False, + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="Seed to determine if node should re-run; actual results are nondeterministic regardless of seed.", + ), + comfy_io.Custom(LumaIO.LUMA_CONCEPTS).Input( + "luma_concepts", + tooltip="Optional Camera Concepts to dictate camera motion via the Luma Concepts node.", + optional=True, + ) + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the video generation", - }, - ), - "model": ([model.value for model in LumaVideoModel],), - "aspect_ratio": ( - [ratio.value for ratio in LumaAspectRatio], - { - "default": LumaAspectRatio.ratio_16_9, - }, - ), - "resolution": ( - [resolution.value for resolution in LumaVideoOutputResolution], - { - "default": LumaVideoOutputResolution.res_540p, - }, - ), - "duration": ([dur.value for dur in LumaVideoModelOutputDuration],), - "loop": ( - IO.BOOLEAN, - { - "default": False, - }, - ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "Seed to determine if node should re-run; actual results are nondeterministic regardless of seed.", - }, - ), - }, - "optional": { - "luma_concepts": ( - LumaIO.LUMA_CONCEPTS, - { - "tooltip": "Optional Camera Concepts to dictate camera motion via the Luma Concepts node." - }, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - async def api_call( - self, + async def execute( + cls, prompt: str, model: str, aspect_ratio: str, @@ -506,13 +520,15 @@ class LumaTextToVideoGenerationNode(ComfyNodeABC): loop: bool, seed, luma_concepts: LumaConceptChain = None, - unique_id: str = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: validate_string(prompt, strip_whitespace=False, min_length=3) duration = duration if model != LumaVideoModel.ray_1_6 else None resolution = resolution if model != LumaVideoModel.ray_1_6 else None + auth_kwargs = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } operation = SynchronousOperation( endpoint=ApiEndpoint( path="/proxy/luma/generations", @@ -529,12 +545,12 @@ class LumaTextToVideoGenerationNode(ComfyNodeABC): loop=loop, concepts=luma_concepts.create_api_model() if luma_concepts else None, ), - auth_kwargs=kwargs, + auth_kwargs=auth_kwargs, ) response_api: LumaGeneration = await operation.execute() - if unique_id: - PromptServer.instance.send_progress_text(f"Luma video generation started: {response_api.id}", unique_id) + if cls.hidden.unique_id: + PromptServer.instance.send_progress_text(f"Luma video generation started: {response_api.id}", cls.hidden.unique_id) operation = PollingOperation( poll_endpoint=ApiEndpoint( @@ -547,90 +563,94 @@ class LumaTextToVideoGenerationNode(ComfyNodeABC): failed_statuses=[LumaState.failed], status_extractor=lambda x: x.state, result_url_extractor=video_result_url_extractor, - node_id=unique_id, + node_id=cls.hidden.unique_id, estimated_duration=LUMA_T2V_AVERAGE_DURATION, - auth_kwargs=kwargs, + auth_kwargs=auth_kwargs, ) response_poll = await operation.execute() async with aiohttp.ClientSession() as session: async with session.get(response_poll.assets.video) as vid_response: - return (VideoFromFile(BytesIO(await vid_response.content.read())),) + return comfy_io.NodeOutput(VideoFromFile(BytesIO(await vid_response.content.read()))) -class LumaImageToVideoGenerationNode(ComfyNodeABC): +class LumaImageToVideoGenerationNode(comfy_io.ComfyNode): """ Generates videos synchronously based on prompt, input images, and output_size. """ - RETURN_TYPES = (IO.VIDEO,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/video/Luma" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="LumaImageToVideoNode", + display_name="Luma Image to Video", + category="api node/video/Luma", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the video generation", + ), + comfy_io.Combo.Input( + "model", + options=[model.value for model in LumaVideoModel], + ), + # comfy_io.Combo.Input( + # "aspect_ratio", + # options=[ratio.value for ratio in LumaAspectRatio], + # default=LumaAspectRatio.ratio_16_9, + # ), + comfy_io.Combo.Input( + "resolution", + options=[resolution.value for resolution in LumaVideoOutputResolution], + default=LumaVideoOutputResolution.res_540p, + ), + comfy_io.Combo.Input( + "duration", + options=[dur.value for dur in LumaVideoModelOutputDuration], + ), + comfy_io.Boolean.Input( + "loop", + default=False, + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="Seed to determine if node should re-run; actual results are nondeterministic regardless of seed.", + ), + comfy_io.Image.Input( + "first_image", + tooltip="First frame of generated video.", + optional=True, + ), + comfy_io.Image.Input( + "last_image", + tooltip="Last frame of generated video.", + optional=True, + ), + comfy_io.Custom(LumaIO.LUMA_CONCEPTS).Input( + "luma_concepts", + tooltip="Optional Camera Concepts to dictate camera motion via the Luma Concepts node.", + optional=True, + ) + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the video generation", - }, - ), - "model": ([model.value for model in LumaVideoModel],), - # "aspect_ratio": ([ratio.value for ratio in LumaAspectRatio], { - # "default": LumaAspectRatio.ratio_16_9, - # }), - "resolution": ( - [resolution.value for resolution in LumaVideoOutputResolution], - { - "default": LumaVideoOutputResolution.res_540p, - }, - ), - "duration": ([dur.value for dur in LumaVideoModelOutputDuration],), - "loop": ( - IO.BOOLEAN, - { - "default": False, - }, - ), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 0xFFFFFFFFFFFFFFFF, - "control_after_generate": True, - "tooltip": "Seed to determine if node should re-run; actual results are nondeterministic regardless of seed.", - }, - ), - }, - "optional": { - "first_image": ( - IO.IMAGE, - {"tooltip": "First frame of generated video."}, - ), - "last_image": (IO.IMAGE, {"tooltip": "Last frame of generated video."}), - "luma_concepts": ( - LumaIO.LUMA_CONCEPTS, - { - "tooltip": "Optional Camera Concepts to dictate camera motion via the Luma Concepts node." - }, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - async def api_call( - self, + async def execute( + cls, prompt: str, model: str, resolution: str, @@ -640,14 +660,16 @@ class LumaImageToVideoGenerationNode(ComfyNodeABC): first_image: torch.Tensor = None, last_image: torch.Tensor = None, luma_concepts: LumaConceptChain = None, - unique_id: str = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: if first_image is None and last_image is None: raise Exception( "At least one of first_image and last_image requires an input." ) - keyframes = await self._convert_to_keyframes(first_image, last_image, auth_kwargs=kwargs) + auth_kwargs = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } + keyframes = await cls._convert_to_keyframes(first_image, last_image, auth_kwargs=auth_kwargs) duration = duration if model != LumaVideoModel.ray_1_6 else None resolution = resolution if model != LumaVideoModel.ray_1_6 else None @@ -668,12 +690,12 @@ class LumaImageToVideoGenerationNode(ComfyNodeABC): keyframes=keyframes, concepts=luma_concepts.create_api_model() if luma_concepts else None, ), - auth_kwargs=kwargs, + auth_kwargs=auth_kwargs, ) response_api: LumaGeneration = await operation.execute() - if unique_id: - PromptServer.instance.send_progress_text(f"Luma video generation started: {response_api.id}", unique_id) + if cls.hidden.unique_id: + PromptServer.instance.send_progress_text(f"Luma video generation started: {response_api.id}", cls.hidden.unique_id) operation = PollingOperation( poll_endpoint=ApiEndpoint( @@ -686,18 +708,19 @@ class LumaImageToVideoGenerationNode(ComfyNodeABC): failed_statuses=[LumaState.failed], status_extractor=lambda x: x.state, result_url_extractor=video_result_url_extractor, - node_id=unique_id, + node_id=cls.hidden.unique_id, estimated_duration=LUMA_I2V_AVERAGE_DURATION, - auth_kwargs=kwargs, + auth_kwargs=auth_kwargs, ) response_poll = await operation.execute() async with aiohttp.ClientSession() as session: async with session.get(response_poll.assets.video) as vid_response: - return (VideoFromFile(BytesIO(await vid_response.content.read())),) + return comfy_io.NodeOutput(VideoFromFile(BytesIO(await vid_response.content.read()))) + @classmethod async def _convert_to_keyframes( - self, + cls, first_image: torch.Tensor = None, last_image: torch.Tensor = None, auth_kwargs: Optional[dict[str,str]] = None, @@ -719,23 +742,18 @@ class LumaImageToVideoGenerationNode(ComfyNodeABC): return LumaKeyframes(frame0=frame0, frame1=frame1) -# A dictionary that contains all nodes you want to export with their names -# NOTE: names should be globally unique -NODE_CLASS_MAPPINGS = { - "LumaImageNode": LumaImageGenerationNode, - "LumaImageModifyNode": LumaImageModifyNode, - "LumaVideoNode": LumaTextToVideoGenerationNode, - "LumaImageToVideoNode": LumaImageToVideoGenerationNode, - "LumaReferenceNode": LumaReferenceNode, - "LumaConceptsNode": LumaConceptsNode, -} +class LumaExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: + return [ + LumaImageGenerationNode, + LumaImageModifyNode, + LumaTextToVideoGenerationNode, + LumaImageToVideoGenerationNode, + LumaReferenceNode, + LumaConceptsNode, + ] -# A dictionary that contains the friendly/humanly readable titles for the nodes -NODE_DISPLAY_NAME_MAPPINGS = { - "LumaImageNode": "Luma Text to Image", - "LumaImageModifyNode": "Luma Image to Image", - "LumaVideoNode": "Luma Text to Video", - "LumaImageToVideoNode": "Luma Image to Video", - "LumaReferenceNode": "Luma Reference", - "LumaConceptsNode": "Luma Concepts", -} + +async def comfy_entrypoint() -> LumaExtension: + return LumaExtension() From ad5aef2d0c8517e971129db1dfb0d0108d8341a8 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 12:34:32 +0300 Subject: [PATCH 033/114] convert nodes_pixart.py to V3 schema (#10019) --- comfy_extras/nodes_pixart.py | 52 +++++++++++++++++++++++------------- 1 file changed, 33 insertions(+), 19 deletions(-) diff --git a/comfy_extras/nodes_pixart.py b/comfy_extras/nodes_pixart.py index 8d9276afe..a23e87b1f 100644 --- a/comfy_extras/nodes_pixart.py +++ b/comfy_extras/nodes_pixart.py @@ -1,24 +1,38 @@ -from nodes import MAX_RESOLUTION +from typing_extensions import override +import nodes +from comfy_api.latest import ComfyExtension, io -class CLIPTextEncodePixArtAlpha: +class CLIPTextEncodePixArtAlpha(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { - "width": ("INT", {"default": 1024.0, "min": 0, "max": MAX_RESOLUTION}), - "height": ("INT", {"default": 1024.0, "min": 0, "max": MAX_RESOLUTION}), - # "aspect_ratio": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01}), - "text": ("STRING", {"multiline": True, "dynamicPrompts": True}), "clip": ("CLIP", ), - }} + def define_schema(cls): + return io.Schema( + node_id="CLIPTextEncodePixArtAlpha", + category="advanced/conditioning", + description="Encodes text and sets the resolution conditioning for PixArt Alpha. Does not apply to PixArt Sigma.", + inputs=[ + io.Int.Input("width", default=1024, min=0, max=nodes.MAX_RESOLUTION), + io.Int.Input("height", default=1024, min=0, max=nodes.MAX_RESOLUTION), + # "aspect_ratio": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01}), + io.String.Input("text", multiline=True, dynamic_prompts=True), + io.Clip.Input("clip"), + ], + outputs=[ + io.Conditioning.Output(), + ], + ) - RETURN_TYPES = ("CONDITIONING",) - FUNCTION = "encode" - CATEGORY = "advanced/conditioning" - DESCRIPTION = "Encodes text and sets the resolution conditioning for PixArt Alpha. Does not apply to PixArt Sigma." - - def encode(self, clip, width, height, text): + @classmethod + def execute(cls, clip, width, height, text): tokens = clip.tokenize(text) - return (clip.encode_from_tokens_scheduled(tokens, add_dict={"width": width, "height": height}),) + return io.NodeOutput(clip.encode_from_tokens_scheduled(tokens, add_dict={"width": width, "height": height})) -NODE_CLASS_MAPPINGS = { - "CLIPTextEncodePixArtAlpha": CLIPTextEncodePixArtAlpha, -} + +class PixArtExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + CLIPTextEncodePixArtAlpha, + ] + +async def comfy_entrypoint() -> PixArtExtension: + return PixArtExtension() From 7eca95657cf7a70c15d598c969b890a164a300a1 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 12:36:43 +0300 Subject: [PATCH 034/114] convert nodes_photomaker.py to V3 schema (#10017) --- comfy_extras/nodes_photomaker.py | 74 ++++++++++++++++++++------------ 1 file changed, 46 insertions(+), 28 deletions(-) diff --git a/comfy_extras/nodes_photomaker.py b/comfy_extras/nodes_photomaker.py index d358ed6d5..228183c07 100644 --- a/comfy_extras/nodes_photomaker.py +++ b/comfy_extras/nodes_photomaker.py @@ -4,6 +4,8 @@ import folder_paths import comfy.clip_model import comfy.clip_vision import comfy.ops +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io # code for model from: https://github.com/TencentARC/PhotoMaker/blob/main/photomaker/model.py under Apache License Version 2.0 VISION_CONFIG_DICT = { @@ -116,41 +118,52 @@ class PhotoMakerIDEncoder(comfy.clip_model.CLIPVisionModelProjection): return updated_prompt_embeds -class PhotoMakerLoader: +class PhotoMakerLoader(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "photomaker_model_name": (folder_paths.get_filename_list("photomaker"), )}} + def define_schema(cls): + return io.Schema( + node_id="PhotoMakerLoader", + category="_for_testing/photomaker", + inputs=[ + io.Combo.Input("photomaker_model_name", options=folder_paths.get_filename_list("photomaker")), + ], + outputs=[ + io.Photomaker.Output(), + ], + is_experimental=True, + ) - RETURN_TYPES = ("PHOTOMAKER",) - FUNCTION = "load_photomaker_model" - - CATEGORY = "_for_testing/photomaker" - - def load_photomaker_model(self, photomaker_model_name): + @classmethod + def execute(cls, photomaker_model_name): photomaker_model_path = folder_paths.get_full_path_or_raise("photomaker", photomaker_model_name) photomaker_model = PhotoMakerIDEncoder() data = comfy.utils.load_torch_file(photomaker_model_path, safe_load=True) if "id_encoder" in data: data = data["id_encoder"] photomaker_model.load_state_dict(data) - return (photomaker_model,) + return io.NodeOutput(photomaker_model) -class PhotoMakerEncode: +class PhotoMakerEncode(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "photomaker": ("PHOTOMAKER",), - "image": ("IMAGE",), - "clip": ("CLIP", ), - "text": ("STRING", {"multiline": True, "dynamicPrompts": True, "default": "photograph of photomaker"}), - }} + def define_schema(cls): + return io.Schema( + node_id="PhotoMakerEncode", + category="_for_testing/photomaker", + inputs=[ + io.Photomaker.Input("photomaker"), + io.Image.Input("image"), + io.Clip.Input("clip"), + io.String.Input("text", multiline=True, dynamic_prompts=True, default="photograph of photomaker"), + ], + outputs=[ + io.Conditioning.Output(), + ], + is_experimental=True, + ) - RETURN_TYPES = ("CONDITIONING",) - FUNCTION = "apply_photomaker" - - CATEGORY = "_for_testing/photomaker" - - def apply_photomaker(self, photomaker, image, clip, text): + @classmethod + def execute(cls, photomaker, image, clip, text): special_token = "photomaker" pixel_values = comfy.clip_vision.clip_preprocess(image.to(photomaker.load_device)).float() try: @@ -178,11 +191,16 @@ class PhotoMakerEncode: else: out = cond - return ([[out, {"pooled_output": pooled}]], ) + return io.NodeOutput([[out, {"pooled_output": pooled}]]) -NODE_CLASS_MAPPINGS = { - "PhotoMakerLoader": PhotoMakerLoader, - "PhotoMakerEncode": PhotoMakerEncode, -} +class PhotomakerExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + PhotoMakerLoader, + PhotoMakerEncode, + ] +async def comfy_entrypoint() -> PhotomakerExtension: + return PhotomakerExtension() From 160698eb418269d64fbbe8c34db27a4d1ddb0540 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 27 Sep 2025 22:25:35 +0300 Subject: [PATCH 035/114] convert nodes_qwen.py to V3 schema (#10049) --- comfy_extras/nodes_qwen.py | 88 ++++++++++++++++++++++---------------- 1 file changed, 51 insertions(+), 37 deletions(-) diff --git a/comfy_extras/nodes_qwen.py b/comfy_extras/nodes_qwen.py index 49747dc7a..525239ae5 100644 --- a/comfy_extras/nodes_qwen.py +++ b/comfy_extras/nodes_qwen.py @@ -1,24 +1,29 @@ import node_helpers import comfy.utils import math +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io -class TextEncodeQwenImageEdit: +class TextEncodeQwenImageEdit(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { - "clip": ("CLIP", ), - "prompt": ("STRING", {"multiline": True, "dynamicPrompts": True}), - }, - "optional": {"vae": ("VAE", ), - "image": ("IMAGE", ),}} + def define_schema(cls): + return io.Schema( + node_id="TextEncodeQwenImageEdit", + category="advanced/conditioning", + inputs=[ + io.Clip.Input("clip"), + io.String.Input("prompt", multiline=True, dynamic_prompts=True), + io.Vae.Input("vae", optional=True), + io.Image.Input("image", optional=True), + ], + outputs=[ + io.Conditioning.Output(), + ], + ) - RETURN_TYPES = ("CONDITIONING",) - FUNCTION = "encode" - - CATEGORY = "advanced/conditioning" - - def encode(self, clip, prompt, vae=None, image=None): + @classmethod + def execute(cls, clip, prompt, vae=None, image=None) -> io.NodeOutput: ref_latent = None if image is None: images = [] @@ -40,28 +45,30 @@ class TextEncodeQwenImageEdit: conditioning = clip.encode_from_tokens_scheduled(tokens) if ref_latent is not None: conditioning = node_helpers.conditioning_set_values(conditioning, {"reference_latents": [ref_latent]}, append=True) - return (conditioning, ) + return io.NodeOutput(conditioning) -class TextEncodeQwenImageEditPlus: +class TextEncodeQwenImageEditPlus(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { - "clip": ("CLIP", ), - "prompt": ("STRING", {"multiline": True, "dynamicPrompts": True}), - }, - "optional": {"vae": ("VAE", ), - "image1": ("IMAGE", ), - "image2": ("IMAGE", ), - "image3": ("IMAGE", ), - }} + def define_schema(cls): + return io.Schema( + node_id="TextEncodeQwenImageEditPlus", + category="advanced/conditioning", + inputs=[ + io.Clip.Input("clip"), + io.String.Input("prompt", multiline=True, dynamic_prompts=True), + io.Vae.Input("vae", optional=True), + io.Image.Input("image1", optional=True), + io.Image.Input("image2", optional=True), + io.Image.Input("image3", optional=True), + ], + outputs=[ + io.Conditioning.Output(), + ], + ) - RETURN_TYPES = ("CONDITIONING",) - FUNCTION = "encode" - - CATEGORY = "advanced/conditioning" - - def encode(self, clip, prompt, vae=None, image1=None, image2=None, image3=None): + @classmethod + def execute(cls, clip, prompt, vae=None, image1=None, image2=None, image3=None) -> io.NodeOutput: ref_latents = [] images = [image1, image2, image3] images_vl = [] @@ -94,10 +101,17 @@ class TextEncodeQwenImageEditPlus: conditioning = clip.encode_from_tokens_scheduled(tokens) if len(ref_latents) > 0: conditioning = node_helpers.conditioning_set_values(conditioning, {"reference_latents": ref_latents}, append=True) - return (conditioning, ) + return io.NodeOutput(conditioning) -NODE_CLASS_MAPPINGS = { - "TextEncodeQwenImageEdit": TextEncodeQwenImageEdit, - "TextEncodeQwenImageEditPlus": TextEncodeQwenImageEditPlus, -} +class QwenExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + TextEncodeQwenImageEdit, + TextEncodeQwenImageEditPlus, + ] + + +async def comfy_entrypoint() -> QwenExtension: + return QwenExtension() From 653ceab4148a9fbc050ebceb674acef760792b77 Mon Sep 17 00:00:00 2001 From: rattus128 <46076784+rattus128@users.noreply.github.com> Date: Sun, 28 Sep 2025 08:14:16 +1000 Subject: [PATCH 036/114] Reduce Peak WAN inference VRAM usage - part II (#10062) * flux: math: Use _addcmul to avoid expensive VRAM intermediate The rope process can be the VRAM peak and this intermediate for the addition result before releasing the original can OOM. addcmul_ it. * wan: Delete the self attention before cross attention This saves VRAM when the cross attention and FFN are in play as the VRAM peak. --- comfy/ldm/flux/math.py | 5 ++++- comfy/ldm/wan/model.py | 1 + 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/comfy/ldm/flux/math.py b/comfy/ldm/flux/math.py index fb7cd7586..8deda0d4a 100644 --- a/comfy/ldm/flux/math.py +++ b/comfy/ldm/flux/math.py @@ -37,7 +37,10 @@ def rope(pos: Tensor, dim: int, theta: int) -> Tensor: def apply_rope1(x: Tensor, freqs_cis: Tensor): x_ = x.to(dtype=freqs_cis.dtype).reshape(*x.shape[:-1], -1, 1, 2) - x_out = freqs_cis[..., 0] * x_[..., 0] + freqs_cis[..., 1] * x_[..., 1] + + x_out = freqs_cis[..., 0] * x_[..., 0] + x_out.addcmul_(freqs_cis[..., 1], x_[..., 1]) + return x_out.reshape(*x.shape).type_as(x) def apply_rope(xq: Tensor, xk: Tensor, freqs_cis: Tensor): diff --git a/comfy/ldm/wan/model.py b/comfy/ldm/wan/model.py index 54616e6eb..0dc650ced 100644 --- a/comfy/ldm/wan/model.py +++ b/comfy/ldm/wan/model.py @@ -237,6 +237,7 @@ class WanAttentionBlock(nn.Module): freqs, transformer_options=transformer_options) x = torch.addcmul(x, y, repeat_e(e[2], x)) + del y # cross-attention & ffn x = x + self.cross_attn(self.norm3(x), context, context_img_len=context_img_len, transformer_options=transformer_options) From 40ae495ddcbc04846e91ccad3e844bb34d98c6fd Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sat, 27 Sep 2025 17:28:49 -0700 Subject: [PATCH 037/114] Improvements to the stable release workflow. (#10065) --- .github/workflows/stable-release.yml | 39 ++++++++++++------- .../windows_release_dependencies.yml | 3 +- 2 files changed, 27 insertions(+), 15 deletions(-) diff --git a/.github/workflows/stable-release.yml b/.github/workflows/stable-release.yml index 2bc8e5905..b39b42acd 100644 --- a/.github/workflows/stable-release.yml +++ b/.github/workflows/stable-release.yml @@ -8,11 +8,11 @@ on: description: 'Git tag' required: true type: string - cu: - description: 'CUDA version' + cache_tag: + description: 'Cached dependencies tag' required: true type: string - default: "129" + default: "cu129" python_minor: description: 'Python minor version' required: true @@ -23,7 +23,11 @@ on: required: true type: string default: "6" - + rel_name: + description: 'Release name' + required: true + type: string + default: "nvidia" jobs: package_comfy_windows: @@ -42,15 +46,15 @@ jobs: id: cache with: path: | - cu${{ inputs.cu }}_python_deps.tar + ${{ inputs.cache_tag }}_python_deps.tar update_comfyui_and_python_dependencies.bat - key: ${{ runner.os }}-build-cu${{ inputs.cu }}-${{ inputs.python_minor }} + key: ${{ runner.os }}-build-${{ inputs.cache_tag }}-${{ inputs.python_minor }} - shell: bash run: | - mv cu${{ inputs.cu }}_python_deps.tar ../ + mv ${{ inputs.cache_tag }}_python_deps.tar ../ mv update_comfyui_and_python_dependencies.bat ../ cd .. - tar xf cu${{ inputs.cu }}_python_deps.tar + tar xf ${{ inputs.cache_tag }}_python_deps.tar pwd ls @@ -65,12 +69,19 @@ jobs: echo 'import site' >> ./python3${{ inputs.python_minor }}._pth curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py ./python.exe get-pip.py - ./python.exe -s -m pip install ../cu${{ inputs.cu }}_python_deps/* + ./python.exe -s -m pip install ../${{ inputs.cache_tag }}_python_deps/* + + grep comfyui ../ComfyUI/requirements.txt ./requirements_comfyui.txt + ./python.exe -s -m pip install -r requirements_comfyui.txt + rm requirements_comfyui.txt + sed -i '1i../ComfyUI' ./python3${{ inputs.python_minor }}._pth - rm ./Lib/site-packages/torch/lib/dnnl.lib #I don't think this is actually used and I need the space - rm ./Lib/site-packages/torch/lib/libprotoc.lib - rm ./Lib/site-packages/torch/lib/libprotobuf.lib + if test -f ./Lib/site-packages/torch/lib/dnnl.lib; then + rm ./Lib/site-packages/torch/lib/dnnl.lib #I don't think this is actually used and I need the space + rm ./Lib/site-packages/torch/lib/libprotoc.lib + rm ./Lib/site-packages/torch/lib/libprotobuf.lib + fi cd .. @@ -91,7 +102,7 @@ jobs: cd .. "C:\Program Files\7-Zip\7z.exe" a -t7z -m0=lzma2 -mx=9 -mfb=128 -md=768m -ms=on -mf=BCJ2 ComfyUI_windows_portable.7z ComfyUI_windows_portable - mv ComfyUI_windows_portable.7z ComfyUI/ComfyUI_windows_portable_nvidia.7z + mv ComfyUI_windows_portable.7z ComfyUI/ComfyUI_windows_portable_${{ inputs.rel_name }}.7z cd ComfyUI_windows_portable python_embeded/python.exe -s ComfyUI/main.py --quick-test-for-ci --cpu @@ -104,7 +115,7 @@ jobs: uses: svenstaro/upload-release-action@v2 with: repo_token: ${{ secrets.GITHUB_TOKEN }} - file: ComfyUI_windows_portable_nvidia.7z + file: ComfyUI_windows_portable_${{ inputs.rel_name }}.7z tag: ${{ inputs.git_tag }} overwrite: true draft: true diff --git a/.github/workflows/windows_release_dependencies.yml b/.github/workflows/windows_release_dependencies.yml index 7761cc1ed..f1e2946e6 100644 --- a/.github/workflows/windows_release_dependencies.yml +++ b/.github/workflows/windows_release_dependencies.yml @@ -56,7 +56,8 @@ jobs: ..\python_embeded\python.exe -s -m pip install --upgrade torch torchvision torchaudio ${{ inputs.xformers }} --extra-index-url https://download.pytorch.org/whl/cu${{ inputs.cu }} -r ../ComfyUI/requirements.txt pygit2 pause" > update_comfyui_and_python_dependencies.bat - python -m pip wheel --no-cache-dir torch torchvision torchaudio ${{ inputs.xformers }} ${{ inputs.extra_dependencies }} --extra-index-url https://download.pytorch.org/whl/cu${{ inputs.cu }} -r requirements.txt pygit2 -w ./temp_wheel_dir + grep -v comfyui requirements.txt > requirements_nocomfyui.txt + python -m pip wheel --no-cache-dir torch torchvision torchaudio ${{ inputs.xformers }} ${{ inputs.extra_dependencies }} --extra-index-url https://download.pytorch.org/whl/cu${{ inputs.cu }} -r requirements_nocomfyui.txt pygit2 -w ./temp_wheel_dir python -m pip install --no-cache-dir ./temp_wheel_dir/* echo installed basic ls -lah temp_wheel_dir From 896f2e653c02769371e113906d70a24306d87a58 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sat, 27 Sep 2025 18:30:35 -0700 Subject: [PATCH 038/114] Fix typo in release workflow. (#10066) --- .github/workflows/stable-release.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/stable-release.yml b/.github/workflows/stable-release.yml index b39b42acd..619b0e995 100644 --- a/.github/workflows/stable-release.yml +++ b/.github/workflows/stable-release.yml @@ -71,7 +71,7 @@ jobs: ./python.exe get-pip.py ./python.exe -s -m pip install ../${{ inputs.cache_tag }}_python_deps/* - grep comfyui ../ComfyUI/requirements.txt ./requirements_comfyui.txt + grep comfyui ../ComfyUI/requirements.txt > ./requirements_comfyui.txt ./python.exe -s -m pip install -r requirements_comfyui.txt rm requirements_comfyui.txt From a1127b232d221432be065f8e765f3538e62a2f41 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sun, 28 Sep 2025 05:11:36 +0300 Subject: [PATCH 039/114] convert nodes_lotus.py to V3 schema (#10057) --- comfy_extras/nodes_lotus.py | 42 +++++++++++++++++++++++-------------- 1 file changed, 26 insertions(+), 16 deletions(-) diff --git a/comfy_extras/nodes_lotus.py b/comfy_extras/nodes_lotus.py index 739dbdd3d..9f62ba2bf 100644 --- a/comfy_extras/nodes_lotus.py +++ b/comfy_extras/nodes_lotus.py @@ -1,20 +1,22 @@ +from typing_extensions import override + import torch import comfy.model_management as mm +from comfy_api.latest import ComfyExtension, io -class LotusConditioning: + +class LotusConditioning(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return { - "required": { - }, - } + def define_schema(cls): + return io.Schema( + node_id="LotusConditioning", + category="conditioning/lotus", + inputs=[], + outputs=[io.Conditioning.Output(display_name="conditioning")], + ) - RETURN_TYPES = ("CONDITIONING",) - RETURN_NAMES = ("conditioning",) - FUNCTION = "conditioning" - CATEGORY = "conditioning/lotus" - - def conditioning(self): + @classmethod + def execute(cls) -> io.NodeOutput: device = mm.get_torch_device() #lotus uses a frozen encoder and null conditioning, i'm just inlining the results of that operation since it doesn't change #and getting parity with the reference implementation would otherwise require inference and 800mb of tensors @@ -22,8 +24,16 @@ class LotusConditioning: cond = [[prompt_embeds, {}]] - return (cond,) + return io.NodeOutput(cond) -NODE_CLASS_MAPPINGS = { - "LotusConditioning" : LotusConditioning, -} + +class LotusExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + LotusConditioning, + ] + + +async def comfy_entrypoint() -> LotusExtension: + return LotusExtension() From 1cf86f5ae5706ff141f8d51ed9ba96ecdcdcb695 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sun, 28 Sep 2025 05:12:51 +0300 Subject: [PATCH 040/114] convert nodes_lumina2.py to V3 schema (#10058) --- comfy_extras/nodes_lumina2.py | 99 +++++++++++++++++++++-------------- 1 file changed, 61 insertions(+), 38 deletions(-) diff --git a/comfy_extras/nodes_lumina2.py b/comfy_extras/nodes_lumina2.py index 275189785..89ff2397a 100644 --- a/comfy_extras/nodes_lumina2.py +++ b/comfy_extras/nodes_lumina2.py @@ -1,20 +1,27 @@ -from comfy.comfy_types import IO, ComfyNodeABC, InputTypeDict +from typing_extensions import override import torch +from comfy_api.latest import ComfyExtension, io -class RenormCFG: + +class RenormCFG(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "model": ("MODEL",), - "cfg_trunc": ("FLOAT", {"default": 100, "min": 0.0, "max": 100.0, "step": 0.01}), - "renorm_cfg": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 100.0, "step": 0.01}), - }} - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" + def define_schema(cls): + return io.Schema( + node_id="RenormCFG", + category="advanced/model", + inputs=[ + io.Model.Input("model"), + io.Float.Input("cfg_trunc", default=100, min=0.0, max=100.0, step=0.01), + io.Float.Input("renorm_cfg", default=1.0, min=0.0, max=100.0, step=0.01), + ], + outputs=[ + io.Model.Output(), + ], + ) - CATEGORY = "advanced/model" - - def patch(self, model, cfg_trunc, renorm_cfg): + @classmethod + def execute(cls, model, cfg_trunc, renorm_cfg) -> io.NodeOutput: def renorm_cfg_func(args): cond_denoised = args["cond_denoised"] uncond_denoised = args["uncond_denoised"] @@ -53,10 +60,10 @@ class RenormCFG: m = model.clone() m.set_model_sampler_cfg_function(renorm_cfg_func) - return (m, ) + return io.NodeOutput(m) -class CLIPTextEncodeLumina2(ComfyNodeABC): +class CLIPTextEncodeLumina2(io.ComfyNode): SYSTEM_PROMPT = { "superior": "You are an assistant designed to generate superior images with the superior "\ "degree of image-text alignment based on textual prompts or user prompts.", @@ -69,36 +76,52 @@ class CLIPTextEncodeLumina2(ComfyNodeABC): "Alignment: You are an assistant designed to generate high-quality images with the highest "\ "degree of image-text alignment based on textual prompts." @classmethod - def INPUT_TYPES(s) -> InputTypeDict: - return { - "required": { - "system_prompt": (list(CLIPTextEncodeLumina2.SYSTEM_PROMPT.keys()), {"tooltip": CLIPTextEncodeLumina2.SYSTEM_PROMPT_TIP}), - "user_prompt": (IO.STRING, {"multiline": True, "dynamicPrompts": True, "tooltip": "The text to be encoded."}), - "clip": (IO.CLIP, {"tooltip": "The CLIP model used for encoding the text."}) - } - } - RETURN_TYPES = (IO.CONDITIONING,) - OUTPUT_TOOLTIPS = ("A conditioning containing the embedded text used to guide the diffusion model.",) - FUNCTION = "encode" + def define_schema(cls): + return io.Schema( + node_id="CLIPTextEncodeLumina2", + display_name="CLIP Text Encode for Lumina2", + category="conditioning", + description="Encodes a system prompt and a user prompt using a CLIP model into an embedding " + "that can be used to guide the diffusion model towards generating specific images.", + inputs=[ + io.Combo.Input( + "system_prompt", + options=list(cls.SYSTEM_PROMPT.keys()), + tooltip=cls.SYSTEM_PROMPT_TIP, + ), + io.String.Input( + "user_prompt", + multiline=True, + dynamic_prompts=True, + tooltip="The text to be encoded.", + ), + io.Clip.Input("clip", tooltip="The CLIP model used for encoding the text."), + ], + outputs=[ + io.Conditioning.Output( + tooltip="A conditioning containing the embedded text used to guide the diffusion model.", + ), + ], + ) - CATEGORY = "conditioning" - DESCRIPTION = "Encodes a system prompt and a user prompt using a CLIP model into an embedding that can be used to guide the diffusion model towards generating specific images." - - def encode(self, clip, user_prompt, system_prompt): + @classmethod + def execute(cls, clip, user_prompt, system_prompt) -> io.NodeOutput: if clip is None: raise RuntimeError("ERROR: clip input is invalid: None\n\nIf the clip is from a checkpoint loader node your checkpoint does not contain a valid clip or text encoder model.") - system_prompt = CLIPTextEncodeLumina2.SYSTEM_PROMPT[system_prompt] + system_prompt = cls.SYSTEM_PROMPT[system_prompt] prompt = f'{system_prompt} {user_prompt}' tokens = clip.tokenize(prompt) - return (clip.encode_from_tokens_scheduled(tokens), ) + return io.NodeOutput(clip.encode_from_tokens_scheduled(tokens)) -NODE_CLASS_MAPPINGS = { - "CLIPTextEncodeLumina2": CLIPTextEncodeLumina2, - "RenormCFG": RenormCFG -} +class Lumina2Extension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + CLIPTextEncodeLumina2, + RenormCFG, + ] -NODE_DISPLAY_NAME_MAPPINGS = { - "CLIPTextEncodeLumina2": "CLIP Text Encode for Lumina2", -} +async def comfy_entrypoint() -> Lumina2Extension: + return Lumina2Extension() From 2dadb348602f8f452eb2a1d8720f6029dc4039a2 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sun, 28 Sep 2025 05:16:22 +0300 Subject: [PATCH 041/114] convert nodes_hypertile.py to V3 schema (#10061) --- comfy_extras/nodes_hypertile.py | 59 +++++++++++++++++++++------------ 1 file changed, 38 insertions(+), 21 deletions(-) diff --git a/comfy_extras/nodes_hypertile.py b/comfy_extras/nodes_hypertile.py index b366117c7..0ad5e6773 100644 --- a/comfy_extras/nodes_hypertile.py +++ b/comfy_extras/nodes_hypertile.py @@ -1,9 +1,11 @@ #Taken from: https://github.com/tfernd/HyperTile/ import math +from typing_extensions import override from einops import rearrange # Use torch rng for consistency across generations from torch import randint +from comfy_api.latest import ComfyExtension, io def random_divisor(value: int, min_value: int, /, max_options: int = 1) -> int: min_value = min(min_value, value) @@ -20,25 +22,31 @@ def random_divisor(value: int, min_value: int, /, max_options: int = 1) -> int: return ns[idx] -class HyperTile: +class HyperTile(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "model": ("MODEL",), - "tile_size": ("INT", {"default": 256, "min": 1, "max": 2048}), - "swap_size": ("INT", {"default": 2, "min": 1, "max": 128}), - "max_depth": ("INT", {"default": 0, "min": 0, "max": 10}), - "scale_depth": ("BOOLEAN", {"default": False}), - }} - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" + def define_schema(cls): + return io.Schema( + node_id="HyperTile", + category="model_patches/unet", + inputs=[ + io.Model.Input("model"), + io.Int.Input("tile_size", default=256, min=1, max=2048), + io.Int.Input("swap_size", default=2, min=1, max=128), + io.Int.Input("max_depth", default=0, min=0, max=10), + io.Boolean.Input("scale_depth", default=False), + ], + outputs=[ + io.Model.Output(), + ], + ) - CATEGORY = "model_patches/unet" - - def patch(self, model, tile_size, swap_size, max_depth, scale_depth): + @classmethod + def execute(cls, model, tile_size, swap_size, max_depth, scale_depth) -> io.NodeOutput: latent_tile_size = max(32, tile_size) // 8 - self.temp = None + temp = None def hypertile_in(q, k, v, extra_options): + nonlocal temp model_chans = q.shape[-2] orig_shape = extra_options['original_shape'] apply_to = [] @@ -58,14 +66,15 @@ class HyperTile: if nh * nw > 1: q = rearrange(q, "b (nh h nw w) c -> (b nh nw) (h w) c", h=h // nh, w=w // nw, nh=nh, nw=nw) - self.temp = (nh, nw, h, w) + temp = (nh, nw, h, w) return q, k, v return q, k, v def hypertile_out(out, extra_options): - if self.temp is not None: - nh, nw, h, w = self.temp - self.temp = None + nonlocal temp + if temp is not None: + nh, nw, h, w = temp + temp = None out = rearrange(out, "(b nh nw) hw c -> b nh nw hw c", nh=nh, nw=nw) out = rearrange(out, "b nh nw (h w) c -> b (nh h nw w) c", h=h // nh, w=w // nw) return out @@ -76,6 +85,14 @@ class HyperTile: m.set_model_attn1_output_patch(hypertile_out) return (m, ) -NODE_CLASS_MAPPINGS = { - "HyperTile": HyperTile, -} + +class HyperTileExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + HyperTile, + ] + + +async def comfy_entrypoint() -> HyperTileExtension: + return HyperTileExtension() From 1364548c721a466adcdc60e49ee291b0d4255245 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Rui=20Wang=20=28=E7=8E=8B=E7=91=9E=29?= Date: Sun, 28 Sep 2025 10:36:02 +0800 Subject: [PATCH 042/114] feat: ComfyUI can be run on the specified Ascend NPU (#9663) * feature: Set the Ascend NPU to use a single one * Enable the `--cuda-device` parameter to support both CUDA and Ascend NPUs simultaneously. * Make the code just set the ASCENT_RT_VISIBLE_DEVICES environment variable without any other edits to master branch --------- Co-authored-by: Jedrzej Kosinski --- main.py | 1 + 1 file changed, 1 insertion(+) diff --git a/main.py b/main.py index c33f0e17b..70696fcc3 100644 --- a/main.py +++ b/main.py @@ -127,6 +127,7 @@ if __name__ == "__main__": if args.cuda_device is not None: os.environ['CUDA_VISIBLE_DEVICES'] = str(args.cuda_device) os.environ['HIP_VISIBLE_DEVICES'] = str(args.cuda_device) + os.environ["ASCEND_RT_VISIBLE_DEVICES"] = str(args.cuda_device) logging.info("Set cuda device to: {}".format(args.cuda_device)) if args.oneapi_device_selector is not None: From 555f902fc1ed20e98201f9102172f0fc190c2c42 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sat, 27 Sep 2025 19:43:25 -0700 Subject: [PATCH 043/114] Fix stable workflow creating multiple draft releases. (#10067) --- .github/workflows/stable-release.yml | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/.github/workflows/stable-release.yml b/.github/workflows/stable-release.yml index 619b0e995..924bdec90 100644 --- a/.github/workflows/stable-release.yml +++ b/.github/workflows/stable-release.yml @@ -112,10 +112,9 @@ jobs: ls - name: Upload binaries to release - uses: svenstaro/upload-release-action@v2 + uses: softprops/action-gh-release@v2 with: - repo_token: ${{ secrets.GITHUB_TOKEN }} - file: ComfyUI_windows_portable_${{ inputs.rel_name }}.7z - tag: ${{ inputs.git_tag }} - overwrite: true + files: ComfyUI_windows_portable_${{ inputs.rel_name }}.7z + tag_name: ${{ inputs.git_tag }} draft: true + overwrite_files: true From b60dc316272ba139e06b8a7b2f5f5b622c9afe20 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sun, 28 Sep 2025 10:41:32 -0700 Subject: [PATCH 044/114] Update command to install latest nighly pytorch. (#10085) --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 3f6cfc2ed..5a257687b 100644 --- a/README.md +++ b/README.md @@ -233,7 +233,7 @@ Nvidia users should install stable pytorch using this command: This is the command to install pytorch nightly instead which might have performance improvements. -```pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu129``` +```pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu130``` #### Troubleshooting From 6ec1cfe101206229ff3af5c3d3675b3b92477067 Mon Sep 17 00:00:00 2001 From: Changrz <51637999+WhiteGiven@users.noreply.github.com> Date: Tue, 30 Sep 2025 02:59:12 +0800 Subject: [PATCH 045/114] [Rodin3d api nodes] Updated the name of the save file path (changed from timestamp to UUID). (#10011) * Update savepath name from time to uuid * delete lib --- comfy_api_nodes/nodes_rodin.py | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) diff --git a/comfy_api_nodes/nodes_rodin.py b/comfy_api_nodes/nodes_rodin.py index 1af393eba..817efb0f5 100644 --- a/comfy_api_nodes/nodes_rodin.py +++ b/comfy_api_nodes/nodes_rodin.py @@ -11,7 +11,6 @@ from comfy.comfy_types.node_typing import IO import folder_paths as comfy_paths import aiohttp import os -import datetime import asyncio import io import logging @@ -243,8 +242,8 @@ class Rodin3DAPI: return mesh_mode, quality_override - async def download_files(self, url_list): - save_path = os.path.join(comfy_paths.get_output_directory(), "Rodin3D", datetime.datetime.now().strftime("%Y-%m-%d_%H-%M-%S")) + async def download_files(self, url_list, task_uuid): + save_path = os.path.join(comfy_paths.get_output_directory(), f"Rodin3D_{task_uuid}") os.makedirs(save_path, exist_ok=True) model_file_path = None async with aiohttp.ClientSession() as session: @@ -320,7 +319,7 @@ class Rodin3D_Regular(Rodin3DAPI): **kwargs) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list) + model = await self.download_files(download_list, task_uuid) return (model,) @@ -366,7 +365,7 @@ class Rodin3D_Detail(Rodin3DAPI): **kwargs) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list) + model = await self.download_files(download_list, task_uuid) return (model,) @@ -412,7 +411,7 @@ class Rodin3D_Smooth(Rodin3DAPI): **kwargs) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list) + model = await self.download_files(download_list, task_uuid) return (model,) @@ -467,7 +466,7 @@ class Rodin3D_Sketch(Rodin3DAPI): ) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list) + model = await self.download_files(download_list, task_uuid) return (model,) From c8276f8c6bee54b494fd5bec8dfb87ed21a3fa65 Mon Sep 17 00:00:00 2001 From: ComfyUI Wiki Date: Tue, 30 Sep 2025 02:59:42 +0800 Subject: [PATCH 046/114] Update template to 0.1.91 (#10096) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index b3f81e8fa..45d3e1607 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,5 +1,5 @@ comfyui-frontend-package==1.26.13 -comfyui-workflow-templates==0.1.88 +comfyui-workflow-templates==0.1.91 comfyui-embedded-docs==0.2.6 torch torchsde From 05a258efd84bfb00e2618eb9b7937b8fef1e82ed Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Mon, 29 Sep 2025 22:01:04 +0300 Subject: [PATCH 047/114] add WanImageToImageApi node (#10094) --- comfy_api_nodes/nodes_wan.py | 149 ++++++++++++++++++++++++++++++++++- 1 file changed, 148 insertions(+), 1 deletion(-) diff --git a/comfy_api_nodes/nodes_wan.py b/comfy_api_nodes/nodes_wan.py index db5bd41c1..0be5daadb 100644 --- a/comfy_api_nodes/nodes_wan.py +++ b/comfy_api_nodes/nodes_wan.py @@ -28,6 +28,12 @@ class Text2ImageInputField(BaseModel): negative_prompt: Optional[str] = Field(None) +class Image2ImageInputField(BaseModel): + prompt: str = Field(...) + negative_prompt: Optional[str] = Field(None) + images: list[str] = Field(..., min_length=1, max_length=2) + + class Text2VideoInputField(BaseModel): prompt: str = Field(...) negative_prompt: Optional[str] = Field(None) @@ -49,6 +55,13 @@ class Txt2ImageParametersField(BaseModel): watermark: bool = Field(True) +class Image2ImageParametersField(BaseModel): + size: Optional[str] = Field(None) + n: int = Field(1, description="Number of images to generate.") # we support only value=1 + seed: int = Field(..., ge=0, le=2147483647) + watermark: bool = Field(True) + + class Text2VideoParametersField(BaseModel): size: str = Field(...) seed: int = Field(..., ge=0, le=2147483647) @@ -73,6 +86,12 @@ class Text2ImageTaskCreationRequest(BaseModel): parameters: Txt2ImageParametersField = Field(...) +class Image2ImageTaskCreationRequest(BaseModel): + model: str = Field(...) + input: Image2ImageInputField = Field(...) + parameters: Image2ImageParametersField = Field(...) + + class Text2VideoTaskCreationRequest(BaseModel): model: str = Field(...) input: Text2VideoInputField = Field(...) @@ -135,7 +154,12 @@ async def process_task( url: str, request_model: Type[T], response_model: Type[R], - payload: Union[Text2ImageTaskCreationRequest, Text2VideoTaskCreationRequest, Image2VideoTaskCreationRequest], + payload: Union[ + Text2ImageTaskCreationRequest, + Image2ImageTaskCreationRequest, + Text2VideoTaskCreationRequest, + Image2VideoTaskCreationRequest, + ], node_id: str, estimated_duration: int, poll_interval: int, @@ -288,6 +312,128 @@ class WanTextToImageApi(comfy_io.ComfyNode): return comfy_io.NodeOutput(await download_url_to_image_tensor(str(response.output.results[0].url))) +class WanImageToImageApi(comfy_io.ComfyNode): + @classmethod + def define_schema(cls): + return comfy_io.Schema( + node_id="WanImageToImageApi", + display_name="Wan Image to Image", + category="api node/image/Wan", + description="Generates an image from one or two input images and a text prompt. " + "The output image is currently fixed at 1.6 MP; its aspect ratio matches the input image(s).", + inputs=[ + comfy_io.Combo.Input( + "model", + options=["wan2.5-i2i-preview"], + default="wan2.5-i2i-preview", + tooltip="Model to use.", + ), + comfy_io.Image.Input( + "image", + tooltip="Single-image editing or multi-image fusion, maximum 2 images.", + ), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt used to describe the elements and visual features, supports English/Chinese.", + ), + comfy_io.String.Input( + "negative_prompt", + multiline=True, + default="", + tooltip="Negative text prompt to guide what to avoid.", + optional=True, + ), + # redo this later as an optional combo of recommended resolutions + # comfy_io.Int.Input( + # "width", + # default=1280, + # min=384, + # max=1440, + # step=16, + # optional=True, + # ), + # comfy_io.Int.Input( + # "height", + # default=1280, + # min=384, + # max=1440, + # step=16, + # optional=True, + # ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=2147483647, + step=1, + display_mode=comfy_io.NumberDisplay.number, + control_after_generate=True, + tooltip="Seed to use for generation.", + optional=True, + ), + comfy_io.Boolean.Input( + "watermark", + default=True, + tooltip="Whether to add an \"AI generated\" watermark to the result.", + optional=True, + ), + ], + outputs=[ + comfy_io.Image.Output(), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, + model: str, + image: torch.Tensor, + prompt: str, + negative_prompt: str = "", + # width: int = 1024, + # height: int = 1024, + seed: int = 0, + watermark: bool = True, + ): + n_images = get_number_of_images(image) + if n_images not in (1, 2): + raise ValueError(f"Expected 1 or 2 input images, got {n_images}.") + images = [] + for i in image: + images.append("data:image/png;base64," + tensor_to_base64_string(i, total_pixels=4096*4096)) + payload = Image2ImageTaskCreationRequest( + model=model, + input=Image2ImageInputField(prompt=prompt, negative_prompt=negative_prompt, images=images), + parameters=Image2ImageParametersField( + # size=f"{width}*{height}", + seed=seed, + watermark=watermark, + ), + ) + response = await process_task( + { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + "/proxy/wan/api/v1/services/aigc/image2image/image-synthesis", + request_model=Image2ImageTaskCreationRequest, + response_model=ImageTaskStatusResponse, + payload=payload, + node_id=cls.hidden.unique_id, + estimated_duration=42, + poll_interval=3, + ) + return comfy_io.NodeOutput(await download_url_to_image_tensor(str(response.output.results[0].url))) + + class WanTextToVideoApi(comfy_io.ComfyNode): @classmethod def define_schema(cls): @@ -593,6 +739,7 @@ class WanApiExtension(ComfyExtension): async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: return [ WanTextToImageApi, + WanImageToImageApi, WanTextToVideoApi, WanImageToVideoApi, ] From b1111c2062ce35d4292bcd94f27c099a13c619cb Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Mon, 29 Sep 2025 22:03:35 +0300 Subject: [PATCH 048/114] convert nodes_mochi.py to V3 schema (#10069) --- comfy_extras/nodes_mochi.py | 49 +++++++++++++++++++++++++------------ 1 file changed, 33 insertions(+), 16 deletions(-) diff --git a/comfy_extras/nodes_mochi.py b/comfy_extras/nodes_mochi.py index 1c474faa9..d750194fc 100644 --- a/comfy_extras/nodes_mochi.py +++ b/comfy_extras/nodes_mochi.py @@ -1,23 +1,40 @@ -import nodes +from typing_extensions import override import torch import comfy.model_management +import nodes +from comfy_api.latest import ComfyExtension, io -class EmptyMochiLatentVideo: + +class EmptyMochiLatentVideo(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "width": ("INT", {"default": 848, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 16}), - "height": ("INT", {"default": 480, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 16}), - "length": ("INT", {"default": 25, "min": 7, "max": nodes.MAX_RESOLUTION, "step": 6}), - "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096})}} - RETURN_TYPES = ("LATENT",) - FUNCTION = "generate" + def define_schema(cls): + return io.Schema( + node_id="EmptyMochiLatentVideo", + category="latent/video", + inputs=[ + io.Int.Input("width", default=848, min=16, max=nodes.MAX_RESOLUTION, step=16), + io.Int.Input("height", default=480, min=16, max=nodes.MAX_RESOLUTION, step=16), + io.Int.Input("length", default=25, min=7, max=nodes.MAX_RESOLUTION, step=6), + io.Int.Input("batch_size", default=1, min=1, max=4096), + ], + outputs=[ + io.Latent.Output(), + ], + ) - CATEGORY = "latent/video" - - def generate(self, width, height, length, batch_size=1): + @classmethod + def execute(cls, width, height, length, batch_size=1) -> io.NodeOutput: latent = torch.zeros([batch_size, 12, ((length - 1) // 6) + 1, height // 8, width // 8], device=comfy.model_management.intermediate_device()) - return ({"samples":latent}, ) + return io.NodeOutput({"samples": latent}) -NODE_CLASS_MAPPINGS = { - "EmptyMochiLatentVideo": EmptyMochiLatentVideo, -} + +class MochiExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + EmptyMochiLatentVideo, + ] + + +async def comfy_entrypoint() -> MochiExtension: + return MochiExtension() From 041b8824f50e01803637d5e83c3f4edaf628f43a Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Mon, 29 Sep 2025 22:05:28 +0300 Subject: [PATCH 049/114] convert nodes_perpneg.py to V3 schema (#10081) --- comfy_extras/nodes_perpneg.py | 93 +++++++++++++++++++++-------------- 1 file changed, 55 insertions(+), 38 deletions(-) diff --git a/comfy_extras/nodes_perpneg.py b/comfy_extras/nodes_perpneg.py index 89e5eef90..cd068ce9c 100644 --- a/comfy_extras/nodes_perpneg.py +++ b/comfy_extras/nodes_perpneg.py @@ -5,6 +5,9 @@ import comfy.samplers import comfy.utils import node_helpers import math +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io + def perp_neg(x, noise_pred_pos, noise_pred_neg, noise_pred_nocond, neg_scale, cond_scale): pos = noise_pred_pos - noise_pred_nocond @@ -16,20 +19,27 @@ def perp_neg(x, noise_pred_pos, noise_pred_neg, noise_pred_nocond, neg_scale, co return cfg_result #TODO: This node should be removed, it has been replaced with PerpNegGuider -class PerpNeg: +class PerpNeg(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"model": ("MODEL", ), - "empty_conditioning": ("CONDITIONING", ), - "neg_scale": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 100.0, "step": 0.01}), - }} - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" + def define_schema(cls): + return io.Schema( + node_id="PerpNeg", + display_name="Perp-Neg (DEPRECATED by PerpNegGuider)", + category="_for_testing", + inputs=[ + io.Model.Input("model"), + io.Conditioning.Input("empty_conditioning"), + io.Float.Input("neg_scale", default=1.0, min=0.0, max=100.0, step=0.01), + ], + outputs=[ + io.Model.Output(), + ], + is_experimental=True, + is_deprecated=True, + ) - CATEGORY = "_for_testing" - DEPRECATED = True - - def patch(self, model, empty_conditioning, neg_scale): + @classmethod + def execute(cls, model, empty_conditioning, neg_scale) -> io.NodeOutput: m = model.clone() nocond = comfy.sampler_helpers.convert_cond(empty_conditioning) @@ -50,7 +60,7 @@ class PerpNeg: m.set_model_sampler_cfg_function(cfg_function) - return (m, ) + return io.NodeOutput(m) class Guider_PerpNeg(comfy.samplers.CFGGuider): @@ -112,35 +122,42 @@ class Guider_PerpNeg(comfy.samplers.CFGGuider): return cfg_result -class PerpNegGuider: +class PerpNegGuider(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": - {"model": ("MODEL",), - "positive": ("CONDITIONING", ), - "negative": ("CONDITIONING", ), - "empty_conditioning": ("CONDITIONING", ), - "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0, "step":0.1, "round": 0.01}), - "neg_scale": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 100.0, "step": 0.01}), - } - } + def define_schema(cls): + return io.Schema( + node_id="PerpNegGuider", + category="_for_testing", + inputs=[ + io.Model.Input("model"), + io.Conditioning.Input("positive"), + io.Conditioning.Input("negative"), + io.Conditioning.Input("empty_conditioning"), + io.Float.Input("cfg", default=8.0, min=0.0, max=100.0, step=0.1, round=0.01), + io.Float.Input("neg_scale", default=1.0, min=0.0, max=100.0, step=0.01), + ], + outputs=[ + io.Guider.Output(), + ], + is_experimental=True, + ) - RETURN_TYPES = ("GUIDER",) - - FUNCTION = "get_guider" - CATEGORY = "_for_testing" - - def get_guider(self, model, positive, negative, empty_conditioning, cfg, neg_scale): + @classmethod + def execute(cls, model, positive, negative, empty_conditioning, cfg, neg_scale) -> io.NodeOutput: guider = Guider_PerpNeg(model) guider.set_conds(positive, negative, empty_conditioning) guider.set_cfg(cfg, neg_scale) - return (guider,) + return io.NodeOutput(guider) -NODE_CLASS_MAPPINGS = { - "PerpNeg": PerpNeg, - "PerpNegGuider": PerpNegGuider, -} -NODE_DISPLAY_NAME_MAPPINGS = { - "PerpNeg": "Perp-Neg (DEPRECATED by PerpNegGuider)", -} +class PerpNegExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + PerpNeg, + PerpNegGuider, + ] + + +async def comfy_entrypoint() -> PerpNegExtension: + return PerpNegExtension() From ed0f4a609b5e6821f97db5cb1715068c25f78e7b Mon Sep 17 00:00:00 2001 From: Christian Byrne Date: Mon, 29 Sep 2025 12:16:02 -0700 Subject: [PATCH 050/114] dont cache new locale entry points (#10101) --- middleware/cache_middleware.py | 11 ++++++----- tests-unit/server_test/test_cache_control.py | 7 +++++++ 2 files changed, 13 insertions(+), 5 deletions(-) diff --git a/middleware/cache_middleware.py b/middleware/cache_middleware.py index 374ef7934..f02135369 100644 --- a/middleware/cache_middleware.py +++ b/middleware/cache_middleware.py @@ -26,11 +26,12 @@ async def cache_control( """Cache control middleware that sets appropriate cache headers based on file type and response status""" response: web.Response = await handler(request) - if ( - request.path.endswith(".js") - or request.path.endswith(".css") - or request.path.endswith("index.json") - ): + path_filename = request.path.rsplit("/", 1)[-1] + is_entry_point = path_filename.startswith("index") and path_filename.endswith( + ".json" + ) + + if request.path.endswith(".js") or request.path.endswith(".css") or is_entry_point: response.headers.setdefault("Cache-Control", "no-cache") return response diff --git a/tests-unit/server_test/test_cache_control.py b/tests-unit/server_test/test_cache_control.py index 8de59125a..fa68d9408 100644 --- a/tests-unit/server_test/test_cache_control.py +++ b/tests-unit/server_test/test_cache_control.py @@ -48,6 +48,13 @@ CACHE_SCENARIOS = [ "expected_cache": "no-cache", "should_have_header": True, }, + { + "name": "localized_index_json_no_cache", + "path": "/templates/index.zh.json", + "status": 200, + "expected_cache": "no-cache", + "should_have_header": True, + }, # Non-matching files { "name": "html_no_header", From 8accf50908094d9cd39168981fa5394274d25491 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Mon, 29 Sep 2025 22:35:51 +0300 Subject: [PATCH 051/114] convert nodes_mahiro.py to V3 schema (#10070) --- comfy_extras/nodes_mahiro.py | 50 ++++++++++++++++++++++++------------ 1 file changed, 33 insertions(+), 17 deletions(-) diff --git a/comfy_extras/nodes_mahiro.py b/comfy_extras/nodes_mahiro.py index 8fcdfba75..07b3353f4 100644 --- a/comfy_extras/nodes_mahiro.py +++ b/comfy_extras/nodes_mahiro.py @@ -1,17 +1,29 @@ +from typing_extensions import override import torch import torch.nn.functional as F -class Mahiro: +from comfy_api.latest import ComfyExtension, io + + +class Mahiro(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"model": ("MODEL",), - }} - RETURN_TYPES = ("MODEL",) - RETURN_NAMES = ("patched_model",) - FUNCTION = "patch" - CATEGORY = "_for_testing" - DESCRIPTION = "Modify the guidance to scale more on the 'direction' of the positive prompt rather than the difference between the negative prompt." - def patch(self, model): + def define_schema(cls): + return io.Schema( + node_id="Mahiro", + display_name="Mahiro is so cute that she deserves a better guidance function!! (。・ω・。)", + category="_for_testing", + description="Modify the guidance to scale more on the 'direction' of the positive prompt rather than the difference between the negative prompt.", + inputs=[ + io.Model.Input("model"), + ], + outputs=[ + io.Model.Output(display_name="patched_model"), + ], + is_experimental=True, + ) + + @classmethod + def execute(cls, model) -> io.NodeOutput: m = model.clone() def mahiro_normd(args): scale: float = args['cond_scale'] @@ -30,12 +42,16 @@ class Mahiro: wm = (simsc*cfg + (4-simsc)*leap) / 4 return wm m.set_model_sampler_post_cfg_function(mahiro_normd) - return (m, ) + return io.NodeOutput(m) -NODE_CLASS_MAPPINGS = { - "Mahiro": Mahiro -} -NODE_DISPLAY_NAME_MAPPINGS = { - "Mahiro": "Mahiro is so cute that she deserves a better guidance function!! (。・ω・。)", -} +class MahiroExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + Mahiro, + ] + + +async def comfy_entrypoint() -> MahiroExtension: + return MahiroExtension() From 7f38e4c538de2fa38d0539c18577cdd0e5d251c2 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 14:27:52 -0700 Subject: [PATCH 052/114] Add action to create cached deps with manually specified torch. (#10102) --- .../windows_release_dependencies_manual.yml | 64 +++++++++++++++++++ 1 file changed, 64 insertions(+) create mode 100644 .github/workflows/windows_release_dependencies_manual.yml diff --git a/.github/workflows/windows_release_dependencies_manual.yml b/.github/workflows/windows_release_dependencies_manual.yml new file mode 100644 index 000000000..0799feef1 --- /dev/null +++ b/.github/workflows/windows_release_dependencies_manual.yml @@ -0,0 +1,64 @@ +name: "Windows Release dependencies Manual" + +on: + workflow_dispatch: + inputs: + torch_dependencies: + description: 'torch dependencies' + required: false + type: string + default: "torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu128" + cache_tag: + description: 'Cached dependencies tag' + required: true + type: string + default: "cu128" + + python_minor: + description: 'python minor version' + required: true + type: string + default: "12" + + python_patch: + description: 'python patch version' + required: true + type: string + default: "10" + +jobs: + build_dependencies: + runs-on: windows-latest + steps: + - uses: actions/checkout@v4 + - uses: actions/setup-python@v5 + with: + python-version: 3.${{ inputs.python_minor }}.${{ inputs.python_patch }} + + - shell: bash + run: | + echo "@echo off + call update_comfyui.bat nopause + echo - + echo This will try to update pytorch and all python dependencies. + echo - + echo If you just want to update normally, close this and run update_comfyui.bat instead. + echo - + pause + ..\python_embeded\python.exe -s -m pip install --upgrade ${{ inputs.torch_dependencies }} -r ../ComfyUI/requirements.txt pygit2 + pause" > update_comfyui_and_python_dependencies.bat + + grep -v comfyui requirements.txt > requirements_nocomfyui.txt + python -m pip wheel --no-cache-dir ${{ inputs.torch_dependencies }} -r requirements_nocomfyui.txt pygit2 -w ./temp_wheel_dir + python -m pip install --no-cache-dir ./temp_wheel_dir/* + echo installed basic + ls -lah temp_wheel_dir + mv temp_wheel_dir ${{ inputs.cache_tag }}_python_deps + tar cf ${{ inputs.cache_tag }}_python_deps.tar ${{ inputs.cache_tag }}_python_deps + + - uses: actions/cache/save@v4 + with: + path: | + ${{ inputs.cache_tag }}_python_deps.tar + update_comfyui_and_python_dependencies.bat + key: ${{ runner.os }}-build-${{ inputs.cache_tag }}-${{ inputs.python_minor }} From 1673ace19b9d63a8dc0d388aafdb54abf2497892 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 16:08:42 -0700 Subject: [PATCH 053/114] Make the final release test optional in the stable release action. (#10103) --- .github/workflows/stable-release.yml | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/.github/workflows/stable-release.yml b/.github/workflows/stable-release.yml index 924bdec90..5eb4a0783 100644 --- a/.github/workflows/stable-release.yml +++ b/.github/workflows/stable-release.yml @@ -28,6 +28,11 @@ on: required: true type: string default: "nvidia" + test_release: + description: 'Test Release' + required: true + type: boolean + default: true jobs: package_comfy_windows: @@ -104,6 +109,10 @@ jobs: "C:\Program Files\7-Zip\7z.exe" a -t7z -m0=lzma2 -mx=9 -mfb=128 -md=768m -ms=on -mf=BCJ2 ComfyUI_windows_portable.7z ComfyUI_windows_portable mv ComfyUI_windows_portable.7z ComfyUI/ComfyUI_windows_portable_${{ inputs.rel_name }}.7z + - shell: bash + if: ${{ inputs.test_release }} + run: | + cd .. cd ComfyUI_windows_portable python_embeded/python.exe -s ComfyUI/main.py --quick-test-for-ci --cpu From 0db6aabed3942ea71258d25d32dc971a2a2421af Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 16:54:05 -0700 Subject: [PATCH 054/114] Different base files for different release. (#10104) --- .github/workflows/stable-release.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/stable-release.yml b/.github/workflows/stable-release.yml index 5eb4a0783..40e1bc157 100644 --- a/.github/workflows/stable-release.yml +++ b/.github/workflows/stable-release.yml @@ -101,7 +101,7 @@ jobs: mkdir update cp -r ComfyUI/.ci/update_windows/* ./update/ - cp -r ComfyUI/.ci/windows_base_files/* ./ + cp -r ComfyUI/.ci/windows_${{ inputs.rel_name }}_base_files/* ./ cp ../update_comfyui_and_python_dependencies.bat ./update/ cd .. From 375884842314a2234ddc29132b03c741ce81443b Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 16:54:37 -0700 Subject: [PATCH 055/114] Different base files for nvidia and amd portables. (#10105) --- .../run_amd_gpu.bat} | 0 .../README_VERY_IMPORTANT.txt | 0 .../run_cpu.bat | 0 .ci/windows_nvidia_base_files/run_nvidia_gpu.bat | 2 ++ .../run_nvidia_gpu_fast_fp16_accumulation.bat | 0 .github/workflows/windows_release_nightly_pytorch.yml | 2 +- .github/workflows/windows_release_package.yml | 2 +- 7 files changed, 4 insertions(+), 2 deletions(-) rename .ci/{windows_base_files/run_nvidia_gpu.bat => windows_amd_base_files/run_amd_gpu.bat} (100%) rename .ci/{windows_base_files => windows_nvidia_base_files}/README_VERY_IMPORTANT.txt (100%) rename .ci/{windows_base_files => windows_nvidia_base_files}/run_cpu.bat (100%) create mode 100755 .ci/windows_nvidia_base_files/run_nvidia_gpu.bat rename .ci/{windows_base_files => windows_nvidia_base_files}/run_nvidia_gpu_fast_fp16_accumulation.bat (100%) diff --git a/.ci/windows_base_files/run_nvidia_gpu.bat b/.ci/windows_amd_base_files/run_amd_gpu.bat similarity index 100% rename from .ci/windows_base_files/run_nvidia_gpu.bat rename to .ci/windows_amd_base_files/run_amd_gpu.bat diff --git a/.ci/windows_base_files/README_VERY_IMPORTANT.txt b/.ci/windows_nvidia_base_files/README_VERY_IMPORTANT.txt similarity index 100% rename from .ci/windows_base_files/README_VERY_IMPORTANT.txt rename to .ci/windows_nvidia_base_files/README_VERY_IMPORTANT.txt diff --git a/.ci/windows_base_files/run_cpu.bat b/.ci/windows_nvidia_base_files/run_cpu.bat similarity index 100% rename from .ci/windows_base_files/run_cpu.bat rename to .ci/windows_nvidia_base_files/run_cpu.bat diff --git a/.ci/windows_nvidia_base_files/run_nvidia_gpu.bat b/.ci/windows_nvidia_base_files/run_nvidia_gpu.bat new file mode 100755 index 000000000..274d7c948 --- /dev/null +++ b/.ci/windows_nvidia_base_files/run_nvidia_gpu.bat @@ -0,0 +1,2 @@ +.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build +pause diff --git a/.ci/windows_base_files/run_nvidia_gpu_fast_fp16_accumulation.bat b/.ci/windows_nvidia_base_files/run_nvidia_gpu_fast_fp16_accumulation.bat similarity index 100% rename from .ci/windows_base_files/run_nvidia_gpu_fast_fp16_accumulation.bat rename to .ci/windows_nvidia_base_files/run_nvidia_gpu_fast_fp16_accumulation.bat diff --git a/.github/workflows/windows_release_nightly_pytorch.yml b/.github/workflows/windows_release_nightly_pytorch.yml index 5bdc940de..ca1ef71ae 100644 --- a/.github/workflows/windows_release_nightly_pytorch.yml +++ b/.github/workflows/windows_release_nightly_pytorch.yml @@ -68,7 +68,7 @@ jobs: mkdir update cp -r ComfyUI/.ci/update_windows/* ./update/ - cp -r ComfyUI/.ci/windows_base_files/* ./ + cp -r ComfyUI/.ci/windows_nvidia_base_files/* ./ cp -r ComfyUI/.ci/windows_nightly_base_files/* ./ echo "call update_comfyui.bat nopause diff --git a/.github/workflows/windows_release_package.yml b/.github/workflows/windows_release_package.yml index 46375698e..7955325fc 100644 --- a/.github/workflows/windows_release_package.yml +++ b/.github/workflows/windows_release_package.yml @@ -81,7 +81,7 @@ jobs: mkdir update cp -r ComfyUI/.ci/update_windows/* ./update/ - cp -r ComfyUI/.ci/windows_base_files/* ./ + cp -r ComfyUI/.ci/windows_nvidia_base_files/* ./ cp ../update_comfyui_and_python_dependencies.bat ./update/ cd .. From 342cf644ce495dafaa31dd49d42c47c5e242e701 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 17:05:44 -0700 Subject: [PATCH 056/114] Add a way to have different names for stable nvidia portables. (#10106) --- .github/workflows/stable-release.yml | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/.github/workflows/stable-release.yml b/.github/workflows/stable-release.yml index 40e1bc157..1cbbfbf69 100644 --- a/.github/workflows/stable-release.yml +++ b/.github/workflows/stable-release.yml @@ -28,6 +28,11 @@ on: required: true type: string default: "nvidia" + rel_extra_name: + description: 'Release extra name' + required: false + type: string + default: "" test_release: description: 'Test Release' required: true @@ -107,7 +112,7 @@ jobs: cd .. "C:\Program Files\7-Zip\7z.exe" a -t7z -m0=lzma2 -mx=9 -mfb=128 -md=768m -ms=on -mf=BCJ2 ComfyUI_windows_portable.7z ComfyUI_windows_portable - mv ComfyUI_windows_portable.7z ComfyUI/ComfyUI_windows_portable_${{ inputs.rel_name }}.7z + mv ComfyUI_windows_portable.7z ComfyUI/ComfyUI_windows_portable_${{ inputs.rel_name }}${{ inputs.rel_extra_name }}.7z - shell: bash if: ${{ inputs.test_release }} @@ -123,7 +128,7 @@ jobs: - name: Upload binaries to release uses: softprops/action-gh-release@v2 with: - files: ComfyUI_windows_portable_${{ inputs.rel_name }}.7z + files: ComfyUI_windows_portable_${{ inputs.rel_name }}${{ inputs.rel_extra_name }}.7z tag_name: ${{ inputs.git_tag }} draft: true overwrite_files: true From bed4b49d08d80e195cb42d5294037fc6b631942e Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 17:31:15 -0700 Subject: [PATCH 057/114] Add action to do the full stable release. (#10107) --- .github/workflows/release-stable-all.yml | 49 ++++++++++++++++++++++++ 1 file changed, 49 insertions(+) create mode 100644 .github/workflows/release-stable-all.yml diff --git a/.github/workflows/release-stable-all.yml b/.github/workflows/release-stable-all.yml new file mode 100644 index 000000000..aac84d637 --- /dev/null +++ b/.github/workflows/release-stable-all.yml @@ -0,0 +1,49 @@ +name: "Release Stable All Portable Versions" + +on: + workflow_dispatch: + inputs: + git_tag: + description: 'Git tag' + required: true + type: string + +jobs: + release_nvidia_default: + name: "Release NVIDIA Default (cu129)" + uses: ./.github/workflows/stable-release.yml + with: + git_tag: ${{ inputs.git_tag }} + cache_tag: "cu129" + python_minor: "13" + python_patch: "6" + rel_name: "nvidia" + rel_extra_name: "" + test_release: true + secrets: inherit + + release_nvidia_cu128: + name: "Release NVIDIA cu128" + uses: ./.github/workflows/stable-release.yml + with: + git_tag: ${{ inputs.git_tag }} + cache_tag: "cu128" + python_minor: "12" + python_patch: "10" + rel_name: "nvidia" + rel_extra_name: "_cu128" + test_release: true + secrets: inherit + + release_amd_rocm: + name: "Release AMD ROCm 6.4.4" + uses: ./.github/workflows/stable-release.yml + with: + git_tag: ${{ inputs.git_tag }} + cache_tag: "rocm644" + python_minor: "12" + python_patch: "10" + rel_name: "amd" + rel_extra_name: "" + test_release: false + secrets: inherit From 447884b65740d9f4160ef13d55adb49ca111140e Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 17:37:51 -0700 Subject: [PATCH 058/114] Make stable release workflow callable. (#10108) --- .github/workflows/stable-release.yml | 36 ++++++++++++++++++++++++++++ 1 file changed, 36 insertions(+) diff --git a/.github/workflows/stable-release.yml b/.github/workflows/stable-release.yml index 1cbbfbf69..28484a9d1 100644 --- a/.github/workflows/stable-release.yml +++ b/.github/workflows/stable-release.yml @@ -2,6 +2,42 @@ name: "Release Stable Version" on: + workflow_call: + inputs: + git_tag: + description: 'Git tag' + required: true + type: string + cache_tag: + description: 'Cached dependencies tag' + required: true + type: string + default: "cu129" + python_minor: + description: 'Python minor version' + required: true + type: string + default: "13" + python_patch: + description: 'Python patch version' + required: true + type: string + default: "6" + rel_name: + description: 'Release name' + required: true + type: string + default: "nvidia" + rel_extra_name: + description: 'Release extra name' + required: false + type: string + default: "" + test_release: + description: 'Test Release' + required: true + type: boolean + default: true workflow_dispatch: inputs: git_tag: From 414a178fb690ef9998f65419f03ef1a83cf559de Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 20:03:02 -0700 Subject: [PATCH 059/114] Add basic readme for AMD portable. (#10109) --- .../README_VERY_IMPORTANT.txt | 24 +++++++++++++++++++ 1 file changed, 24 insertions(+) create mode 100755 .ci/windows_amd_base_files/README_VERY_IMPORTANT.txt diff --git a/.ci/windows_amd_base_files/README_VERY_IMPORTANT.txt b/.ci/windows_amd_base_files/README_VERY_IMPORTANT.txt new file mode 100755 index 000000000..570ac3398 --- /dev/null +++ b/.ci/windows_amd_base_files/README_VERY_IMPORTANT.txt @@ -0,0 +1,24 @@ +As of the time of writing this you need this preview driver for best results: +https://www.amd.com/en/resources/support-articles/release-notes/RN-AMDGPU-WINDOWS-PYTORCH-PREVIEW.html + +HOW TO RUN: + +if you have a AMD gpu: + +run_amd_gpu.bat + + +IF YOU GET A RED ERROR IN THE UI MAKE SURE YOU HAVE A MODEL/CHECKPOINT IN: ComfyUI\models\checkpoints + +You can download the stable diffusion XL one from: https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/sd_xl_base_1.0_0.9vae.safetensors + + +RECOMMENDED WAY TO UPDATE: +To update the ComfyUI code: update\update_comfyui.bat + + +TO SHARE MODELS BETWEEN COMFYUI AND ANOTHER UI: +In the ComfyUI directory you will find a file: extra_model_paths.yaml.example +Rename this file to: extra_model_paths.yaml and edit it with your favorite text editor. + + From 977a4ed8c55ade53d0d6cfe1fe8a6396ee35a2ec Mon Sep 17 00:00:00 2001 From: comfyanonymous Date: Mon, 29 Sep 2025 23:04:42 -0400 Subject: [PATCH 060/114] ComfyUI version 0.3.61 --- comfyui_version.py | 2 +- pyproject.toml | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/comfyui_version.py b/comfyui_version.py index d469a8194..737b72131 100644 --- a/comfyui_version.py +++ b/comfyui_version.py @@ -1,3 +1,3 @@ # This file is automatically generated by the build process when version is # updated in pyproject.toml. -__version__ = "0.3.60" +__version__ = "0.3.61" diff --git a/pyproject.toml b/pyproject.toml index 7340c320b..e851560f7 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -1,6 +1,6 @@ [project] name = "ComfyUI" -version = "0.3.60" +version = "0.3.61" readme = "README.md" license = { file = "LICENSE" } requires-python = ">=3.9" From 6e079abc3a3fc0fb98e2a0848877874151310ed1 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 29 Sep 2025 20:11:37 -0700 Subject: [PATCH 061/114] Workflow permission fix. (#10110) --- .github/workflows/release-stable-all.yml | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/.github/workflows/release-stable-all.yml b/.github/workflows/release-stable-all.yml index aac84d637..5c1024599 100644 --- a/.github/workflows/release-stable-all.yml +++ b/.github/workflows/release-stable-all.yml @@ -10,6 +10,10 @@ on: jobs: release_nvidia_default: + permissions: + contents: "write" + packages: "write" + pull-requests: "read" name: "Release NVIDIA Default (cu129)" uses: ./.github/workflows/stable-release.yml with: @@ -23,6 +27,10 @@ jobs: secrets: inherit release_nvidia_cu128: + permissions: + contents: "write" + packages: "write" + pull-requests: "read" name: "Release NVIDIA cu128" uses: ./.github/workflows/stable-release.yml with: @@ -36,6 +44,10 @@ jobs: secrets: inherit release_amd_rocm: + permissions: + contents: "write" + packages: "write" + pull-requests: "read" name: "Release AMD ROCm 6.4.4" uses: ./.github/workflows/stable-release.yml with: From f48d7230de2f7b10fe8bfda3d7f53241d19c7266 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Tue, 30 Sep 2025 09:17:49 -0700 Subject: [PATCH 062/114] Add new portable links to readme. (#10112) --- README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/README.md b/README.md index 5a257687b..8f24a33ee 100644 --- a/README.md +++ b/README.md @@ -176,6 +176,12 @@ Simply download, extract with [7-Zip](https://7-zip.org) and run. Make sure you If you have trouble extracting it, right click the file -> properties -> unblock +#### Alternative Downloads: + +[Experimental portable for AMD GPUs](https://github.com/comfyanonymous/ComfyUI/releases/latest/download/ComfyUI_windows_portable_amd.7z) + +[Portable with pytorch cuda 12.8 and python 3.12](https://github.com/comfyanonymous/ComfyUI/releases/latest/download/ComfyUI_windows_portable_nvidia_cu128.7z) (Supports Nvidia 10 series and older GPUs). + #### How do I share models between another UI and ComfyUI? See the [Config file](extra_model_paths.yaml.example) to set the search paths for models. In the standalone windows build you can find this file in the ComfyUI directory. Rename this file to extra_model_paths.yaml and edit it with your favorite text editor. From 631b9ae861bf8bdd3c538da232e4c8938448e59d Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Tue, 30 Sep 2025 20:21:47 +0300 Subject: [PATCH 063/114] fix(Rodin3D-Gen2): missing "task_uuid" parameter (#10128) --- comfy_api_nodes/nodes_rodin.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/comfy_api_nodes/nodes_rodin.py b/comfy_api_nodes/nodes_rodin.py index 817efb0f5..633ac46d3 100644 --- a/comfy_api_nodes/nodes_rodin.py +++ b/comfy_api_nodes/nodes_rodin.py @@ -540,7 +540,7 @@ class Rodin3D_Gen2(Rodin3DAPI): **kwargs) await self.poll_for_task_status(subscription_key, **kwargs) download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list) + model = await self.download_files(download_list, task_uuid) return (model,) From b682a73c55a6434fdd9293d45ace969597f8ad65 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Tue, 30 Sep 2025 20:43:41 +0300 Subject: [PATCH 064/114] enable Seedance Pro model in the FirstLastFrame node (#10120) --- comfy_api_nodes/nodes_bytedance.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/comfy_api_nodes/nodes_bytedance.py b/comfy_api_nodes/nodes_bytedance.py index a7eeaf15a..654d6a362 100644 --- a/comfy_api_nodes/nodes_bytedance.py +++ b/comfy_api_nodes/nodes_bytedance.py @@ -920,7 +920,7 @@ class ByteDanceFirstLastFrameNode(comfy_io.ComfyNode): inputs=[ comfy_io.Combo.Input( "model", - options=[Image2VideoModelName.seedance_1_lite.value], + options=[model.value for model in Image2VideoModelName], default=Image2VideoModelName.seedance_1_lite.value, tooltip="Model name", ), From bab8ba20bf47d985d6b1d73627c2add76bd4e716 Mon Sep 17 00:00:00 2001 From: comfyanonymous Date: Tue, 30 Sep 2025 15:12:07 -0400 Subject: [PATCH 065/114] ComfyUI version 0.3.62. --- comfyui_version.py | 2 +- pyproject.toml | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/comfyui_version.py b/comfyui_version.py index 737b72131..ac76fbe35 100644 --- a/comfyui_version.py +++ b/comfyui_version.py @@ -1,3 +1,3 @@ # This file is automatically generated by the build process when version is # updated in pyproject.toml. -__version__ = "0.3.61" +__version__ = "0.3.62" diff --git a/pyproject.toml b/pyproject.toml index e851560f7..d0a76c6d0 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -1,6 +1,6 @@ [project] name = "ComfyUI" -version = "0.3.61" +version = "0.3.62" readme = "README.md" license = { file = "LICENSE" } requires-python = ">=3.9" From c4a8cf60ab5d6eaf052b7a08f5ee97104acf7a2f Mon Sep 17 00:00:00 2001 From: AustinMroz Date: Tue, 30 Sep 2025 22:12:32 -0700 Subject: [PATCH 066/114] Bump frontend to 1.27.7 (#10133) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 45d3e1607..588c5dcf0 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,4 +1,4 @@ -comfyui-frontend-package==1.26.13 +comfyui-frontend-package==1.27.7 comfyui-workflow-templates==0.1.91 comfyui-embedded-docs==0.2.6 torch From 638097829d2352a1c78ab4fbb1e028d1e7cff012 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 1 Oct 2025 09:00:22 +0300 Subject: [PATCH 067/114] convert nodes_audio_encoder.py to V3 schema (#10123) --- comfy_api/latest/_io.py | 1 + comfy_extras/nodes_audio_encoder.py | 68 ++++++++++++++++++----------- 2 files changed, 44 insertions(+), 25 deletions(-) diff --git a/comfy_api/latest/_io.py b/comfy_api/latest/_io.py index 4826818df..2d95cffd6 100644 --- a/comfy_api/latest/_io.py +++ b/comfy_api/latest/_io.py @@ -1605,6 +1605,7 @@ class _IO: Model = Model ClipVision = ClipVision ClipVisionOutput = ClipVisionOutput + AudioEncoder = AudioEncoder AudioEncoderOutput = AudioEncoderOutput StyleModel = StyleModel Gligen = Gligen diff --git a/comfy_extras/nodes_audio_encoder.py b/comfy_extras/nodes_audio_encoder.py index 39a140fef..13aacd41a 100644 --- a/comfy_extras/nodes_audio_encoder.py +++ b/comfy_extras/nodes_audio_encoder.py @@ -1,44 +1,62 @@ import folder_paths import comfy.audio_encoders.audio_encoders import comfy.utils +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io -class AudioEncoderLoader: +class AudioEncoderLoader(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "audio_encoder_name": (folder_paths.get_filename_list("audio_encoders"), ), - }} - RETURN_TYPES = ("AUDIO_ENCODER",) - FUNCTION = "load_model" + def define_schema(cls) -> io.Schema: + return io.Schema( + node_id="AudioEncoderLoader", + category="loaders", + inputs=[ + io.Combo.Input( + "audio_encoder_name", + options=folder_paths.get_filename_list("audio_encoders"), + ), + ], + outputs=[io.AudioEncoder.Output()], + ) - CATEGORY = "loaders" - - def load_model(self, audio_encoder_name): + @classmethod + def execute(cls, audio_encoder_name) -> io.NodeOutput: audio_encoder_name = folder_paths.get_full_path_or_raise("audio_encoders", audio_encoder_name) sd = comfy.utils.load_torch_file(audio_encoder_name, safe_load=True) audio_encoder = comfy.audio_encoders.audio_encoders.load_audio_encoder_from_sd(sd) if audio_encoder is None: raise RuntimeError("ERROR: audio encoder file is invalid and does not contain a valid model.") - return (audio_encoder,) + return io.NodeOutput(audio_encoder) -class AudioEncoderEncode: +class AudioEncoderEncode(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "audio_encoder": ("AUDIO_ENCODER",), - "audio": ("AUDIO",), - }} - RETURN_TYPES = ("AUDIO_ENCODER_OUTPUT",) - FUNCTION = "encode" + def define_schema(cls) -> io.Schema: + return io.Schema( + node_id="AudioEncoderEncode", + category="conditioning", + inputs=[ + io.AudioEncoder.Input("audio_encoder"), + io.Audio.Input("audio"), + ], + outputs=[io.AudioEncoderOutput.Output()], + ) - CATEGORY = "conditioning" - - def encode(self, audio_encoder, audio): + @classmethod + def execute(cls, audio_encoder, audio) -> io.NodeOutput: output = audio_encoder.encode_audio(audio["waveform"], audio["sample_rate"]) - return (output,) + return io.NodeOutput(output) -NODE_CLASS_MAPPINGS = { - "AudioEncoderLoader": AudioEncoderLoader, - "AudioEncoderEncode": AudioEncoderEncode, -} +class AudioEncoder(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + AudioEncoderLoader, + AudioEncoderEncode, + ] + + +async def comfy_entrypoint() -> AudioEncoder: + return AudioEncoder() From 7eb7160db487feb891ceabdf985b09f9a8091869 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 1 Oct 2025 22:16:59 +0300 Subject: [PATCH 068/114] convert nodes_gits.py to V3 schema (#9949) --- comfy_extras/nodes_gits.py | 49 ++++++++++++++++++++++++-------------- 1 file changed, 31 insertions(+), 18 deletions(-) diff --git a/comfy_extras/nodes_gits.py b/comfy_extras/nodes_gits.py index 47b1dd049..25367560a 100644 --- a/comfy_extras/nodes_gits.py +++ b/comfy_extras/nodes_gits.py @@ -1,6 +1,8 @@ # from https://github.com/zju-pi/diff-sampler/tree/main/gits-main import numpy as np import torch +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io def loglinear_interp(t_steps, num_steps): """ @@ -333,25 +335,28 @@ NOISE_LEVELS = { ], } -class GITSScheduler: +class GITSScheduler(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": - {"coeff": ("FLOAT", {"default": 1.20, "min": 0.80, "max": 1.50, "step": 0.05}), - "steps": ("INT", {"default": 10, "min": 2, "max": 1000}), - "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), - } - } - RETURN_TYPES = ("SIGMAS",) - CATEGORY = "sampling/custom_sampling/schedulers" + def define_schema(cls): + return io.Schema( + node_id="GITSScheduler", + category="sampling/custom_sampling/schedulers", + inputs=[ + io.Float.Input("coeff", default=1.20, min=0.80, max=1.50, step=0.05), + io.Int.Input("steps", default=10, min=2, max=1000), + io.Float.Input("denoise", default=1.0, min=0.0, max=1.0, step=0.01), + ], + outputs=[ + io.Sigmas.Output(), + ], + ) - FUNCTION = "get_sigmas" - - def get_sigmas(self, coeff, steps, denoise): + @classmethod + def execute(cls, coeff, steps, denoise): total_steps = steps if denoise < 1.0: if denoise <= 0.0: - return (torch.FloatTensor([]),) + return io.NodeOutput(torch.FloatTensor([])) total_steps = round(steps * denoise) if steps <= 20: @@ -362,8 +367,16 @@ class GITSScheduler: sigmas = sigmas[-(total_steps + 1):] sigmas[-1] = 0 - return (torch.FloatTensor(sigmas), ) + return io.NodeOutput(torch.FloatTensor(sigmas)) -NODE_CLASS_MAPPINGS = { - "GITSScheduler": GITSScheduler, -} + +class GITSSchedulerExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + GITSScheduler, + ] + + +async def comfy_entrypoint() -> GITSSchedulerExtension: + return GITSSchedulerExtension() From e0210ce0a7140e0c61bce7fdb964b5e5e8d31619 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 1 Oct 2025 22:17:33 +0300 Subject: [PATCH 069/114] convert nodes_differential_diffusion.py to V3 schema (#10056) --- comfy_extras/nodes_differential_diffusion.py | 69 ++++++++++++-------- 1 file changed, 40 insertions(+), 29 deletions(-) diff --git a/comfy_extras/nodes_differential_diffusion.py b/comfy_extras/nodes_differential_diffusion.py index 255ac420d..6dfdf466c 100644 --- a/comfy_extras/nodes_differential_diffusion.py +++ b/comfy_extras/nodes_differential_diffusion.py @@ -1,34 +1,41 @@ # code adapted from https://github.com/exx8/differential-diffusion +from typing_extensions import override + import torch +from comfy_api.latest import ComfyExtension, io -class DifferentialDiffusion(): + +class DifferentialDiffusion(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return { - "required": { - "model": ("MODEL", ), - }, - "optional": { - "strength": ("FLOAT", { - "default": 1.0, - "min": 0.0, - "max": 1.0, - "step": 0.01, - }), - } - } - RETURN_TYPES = ("MODEL",) - FUNCTION = "apply" - CATEGORY = "_for_testing" - INIT = False + def define_schema(cls): + return io.Schema( + node_id="DifferentialDiffusion", + display_name="Differential Diffusion", + category="_for_testing", + inputs=[ + io.Model.Input("model"), + io.Float.Input( + "strength", + default=1.0, + min=0.0, + max=1.0, + step=0.01, + optional=True, + ), + ], + outputs=[io.Model.Output()], + is_experimental=True, + ) - def apply(self, model, strength=1.0): + @classmethod + def execute(cls, model, strength=1.0) -> io.NodeOutput: model = model.clone() - model.set_model_denoise_mask_function(lambda *args, **kwargs: self.forward(*args, **kwargs, strength=strength)) - return (model, ) + model.set_model_denoise_mask_function(lambda *args, **kwargs: cls.forward(*args, **kwargs, strength=strength)) + return io.NodeOutput(model) - def forward(self, sigma: torch.Tensor, denoise_mask: torch.Tensor, extra_options: dict, strength: float): + @classmethod + def forward(cls, sigma: torch.Tensor, denoise_mask: torch.Tensor, extra_options: dict, strength: float): model = extra_options["model"] step_sigmas = extra_options["sigmas"] sigma_to = model.inner_model.model_sampling.sigma_min @@ -53,9 +60,13 @@ class DifferentialDiffusion(): return binary_mask -NODE_CLASS_MAPPINGS = { - "DifferentialDiffusion": DifferentialDiffusion, -} -NODE_DISPLAY_NAME_MAPPINGS = { - "DifferentialDiffusion": "Differential Diffusion", -} +class DifferentialDiffusionExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + DifferentialDiffusion, + ] + + +async def comfy_entrypoint() -> DifferentialDiffusionExtension: + return DifferentialDiffusionExtension() From 3af1881455fb0c44c3030b2d61b79302933386d2 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 1 Oct 2025 22:18:04 +0300 Subject: [PATCH 070/114] convert nodes_optimalsteps.py to V3 schema (#10074) --- comfy_extras/nodes_optimalsteps.py | 52 +++++++++++++++++++----------- 1 file changed, 33 insertions(+), 19 deletions(-) diff --git a/comfy_extras/nodes_optimalsteps.py b/comfy_extras/nodes_optimalsteps.py index e7c851ca2..73f0104d8 100644 --- a/comfy_extras/nodes_optimalsteps.py +++ b/comfy_extras/nodes_optimalsteps.py @@ -1,9 +1,12 @@ # from https://github.com/bebebe666/OptimalSteps - import numpy as np import torch +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io + + def loglinear_interp(t_steps, num_steps): """ Performs log-linear interpolation of a given array of decreasing numbers. @@ -23,25 +26,28 @@ NOISE_LEVELS = {"FLUX": [0.9968, 0.9886, 0.9819, 0.975, 0.966, 0.9471, 0.9158, 0 "Chroma": [0.992, 0.99, 0.988, 0.985, 0.982, 0.978, 0.973, 0.968, 0.961, 0.953, 0.943, 0.931, 0.917, 0.9, 0.881, 0.858, 0.832, 0.802, 0.769, 0.731, 0.69, 0.646, 0.599, 0.55, 0.501, 0.451, 0.402, 0.355, 0.311, 0.27, 0.232, 0.199, 0.169, 0.143, 0.12, 0.101, 0.084, 0.07, 0.058, 0.048, 0.001], } -class OptimalStepsScheduler: +class OptimalStepsScheduler(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": - {"model_type": (["FLUX", "Wan", "Chroma"], ), - "steps": ("INT", {"default": 20, "min": 3, "max": 1000}), - "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), - } - } - RETURN_TYPES = ("SIGMAS",) - CATEGORY = "sampling/custom_sampling/schedulers" + def define_schema(cls): + return io.Schema( + node_id="OptimalStepsScheduler", + category="sampling/custom_sampling/schedulers", + inputs=[ + io.Combo.Input("model_type", options=["FLUX", "Wan", "Chroma"]), + io.Int.Input("steps", default=20, min=3, max=1000), + io.Float.Input("denoise", default=1.0, min=0.0, max=1.0, step=0.01), + ], + outputs=[ + io.Sigmas.Output(), + ], + ) - FUNCTION = "get_sigmas" - - def get_sigmas(self, model_type, steps, denoise): + @classmethod + def execute(cls, model_type, steps, denoise) ->io.NodeOutput: total_steps = steps if denoise < 1.0: if denoise <= 0.0: - return (torch.FloatTensor([]),) + return io.NodeOutput(torch.FloatTensor([])) total_steps = round(steps * denoise) sigmas = NOISE_LEVELS[model_type][:] @@ -50,8 +56,16 @@ class OptimalStepsScheduler: sigmas = sigmas[-(total_steps + 1):] sigmas[-1] = 0 - return (torch.FloatTensor(sigmas), ) + return io.NodeOutput(torch.FloatTensor(sigmas)) -NODE_CLASS_MAPPINGS = { - "OptimalStepsScheduler": OptimalStepsScheduler, -} + +class OptimalStepsExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + OptimalStepsScheduler, + ] + + +async def comfy_entrypoint() -> OptimalStepsExtension: + return OptimalStepsExtension() From 11bab7be76d0bfdb326e8aea53cdfebd99b42cc5 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 1 Oct 2025 22:18:49 +0300 Subject: [PATCH 071/114] convert nodes_pag.py to V3 schema (#10080) --- comfy_extras/nodes_pag.py | 49 +++++++++++++++++++++++++-------------- 1 file changed, 31 insertions(+), 18 deletions(-) diff --git a/comfy_extras/nodes_pag.py b/comfy_extras/nodes_pag.py index eb28196f4..79fea5f0c 100644 --- a/comfy_extras/nodes_pag.py +++ b/comfy_extras/nodes_pag.py @@ -3,25 +3,30 @@ #My modified one here is more basic but has less chances of breaking with ComfyUI updates. +from typing_extensions import override + import comfy.model_patcher import comfy.samplers +from comfy_api.latest import ComfyExtension, io -class PerturbedAttentionGuidance: + +class PerturbedAttentionGuidance(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return { - "required": { - "model": ("MODEL",), - "scale": ("FLOAT", {"default": 3.0, "min": 0.0, "max": 100.0, "step": 0.01, "round": 0.01}), - } - } + def define_schema(cls): + return io.Schema( + node_id="PerturbedAttentionGuidance", + category="model_patches/unet", + inputs=[ + io.Model.Input("model"), + io.Float.Input("scale", default=3.0, min=0.0, max=100.0, step=0.01, round=0.01), + ], + outputs=[ + io.Model.Output(), + ], + ) - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" - - CATEGORY = "model_patches/unet" - - def patch(self, model, scale): + @classmethod + def execute(cls, model, scale) -> io.NodeOutput: unet_block = "middle" unet_block_id = 0 m = model.clone() @@ -49,8 +54,16 @@ class PerturbedAttentionGuidance: m.set_model_sampler_post_cfg_function(post_cfg_function) - return (m,) + return io.NodeOutput(m) -NODE_CLASS_MAPPINGS = { - "PerturbedAttentionGuidance": PerturbedAttentionGuidance, -} + +class PAGExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + PerturbedAttentionGuidance, + ] + + +async def comfy_entrypoint() -> PAGExtension: + return PAGExtension() From d9c0a4053d955c7fd3400be07001bc4e774591e1 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 1 Oct 2025 22:19:56 +0300 Subject: [PATCH 072/114] convert nodes_lt.py to V3 schema (#10084) --- comfy_extras/nodes_lt.py | 412 ++++++++++++++++++++++----------------- 1 file changed, 228 insertions(+), 184 deletions(-) diff --git a/comfy_extras/nodes_lt.py b/comfy_extras/nodes_lt.py index f82337a67..b51d15804 100644 --- a/comfy_extras/nodes_lt.py +++ b/comfy_extras/nodes_lt.py @@ -1,4 +1,3 @@ -import io import nodes import node_helpers import torch @@ -8,46 +7,60 @@ import comfy.utils import math import numpy as np import av +from io import BytesIO +from typing_extensions import override from comfy.ldm.lightricks.symmetric_patchifier import SymmetricPatchifier, latent_to_pixel_coords +from comfy_api.latest import ComfyExtension, io -class EmptyLTXVLatentVideo: +class EmptyLTXVLatentVideo(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "width": ("INT", {"default": 768, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 32}), - "height": ("INT", {"default": 512, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 32}), - "length": ("INT", {"default": 97, "min": 1, "max": nodes.MAX_RESOLUTION, "step": 8}), - "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096})}} - RETURN_TYPES = ("LATENT",) - FUNCTION = "generate" + def define_schema(cls): + return io.Schema( + node_id="EmptyLTXVLatentVideo", + category="latent/video/ltxv", + inputs=[ + io.Int.Input("width", default=768, min=64, max=nodes.MAX_RESOLUTION, step=32), + io.Int.Input("height", default=512, min=64, max=nodes.MAX_RESOLUTION, step=32), + io.Int.Input("length", default=97, min=1, max=nodes.MAX_RESOLUTION, step=8), + io.Int.Input("batch_size", default=1, min=1, max=4096), + ], + outputs=[ + io.Latent.Output(), + ], + ) - CATEGORY = "latent/video/ltxv" - - def generate(self, width, height, length, batch_size=1): + @classmethod + def execute(cls, width, height, length, batch_size=1) -> io.NodeOutput: latent = torch.zeros([batch_size, 128, ((length - 1) // 8) + 1, height // 32, width // 32], device=comfy.model_management.intermediate_device()) - return ({"samples": latent}, ) + return io.NodeOutput({"samples": latent}) -class LTXVImgToVideo: +class LTXVImgToVideo(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"positive": ("CONDITIONING", ), - "negative": ("CONDITIONING", ), - "vae": ("VAE",), - "image": ("IMAGE",), - "width": ("INT", {"default": 768, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 32}), - "height": ("INT", {"default": 512, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 32}), - "length": ("INT", {"default": 97, "min": 9, "max": nodes.MAX_RESOLUTION, "step": 8}), - "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096}), - "strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0}), - }} + def define_schema(cls): + return io.Schema( + node_id="LTXVImgToVideo", + category="conditioning/video_models", + inputs=[ + io.Conditioning.Input("positive"), + io.Conditioning.Input("negative"), + io.Vae.Input("vae"), + io.Image.Input("image"), + io.Int.Input("width", default=768, min=64, max=nodes.MAX_RESOLUTION, step=32), + io.Int.Input("height", default=512, min=64, max=nodes.MAX_RESOLUTION, step=32), + io.Int.Input("length", default=97, min=9, max=nodes.MAX_RESOLUTION, step=8), + io.Int.Input("batch_size", default=1, min=1, max=4096), + io.Float.Input("strength", default=1.0, min=0.0, max=1.0), + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + io.Latent.Output(display_name="latent"), + ], + ) - RETURN_TYPES = ("CONDITIONING", "CONDITIONING", "LATENT") - RETURN_NAMES = ("positive", "negative", "latent") - - CATEGORY = "conditioning/video_models" - FUNCTION = "generate" - - def generate(self, positive, negative, image, vae, width, height, length, batch_size, strength): + @classmethod + def execute(cls, positive, negative, image, vae, width, height, length, batch_size, strength) -> io.NodeOutput: pixels = comfy.utils.common_upscale(image.movedim(-1, 1), width, height, "bilinear", "center").movedim(1, -1) encode_pixels = pixels[:, :, :, :3] t = vae.encode(encode_pixels) @@ -62,7 +75,7 @@ class LTXVImgToVideo: ) conditioning_latent_frames_mask[:, :, :t.shape[2]] = 1.0 - strength - return (positive, negative, {"samples": latent, "noise_mask": conditioning_latent_frames_mask}, ) + return io.NodeOutput(positive, negative, {"samples": latent, "noise_mask": conditioning_latent_frames_mask}) def conditioning_get_any_value(conditioning, key, default=None): @@ -93,35 +106,46 @@ def get_keyframe_idxs(cond): num_keyframes = torch.unique(keyframe_idxs[:, 0]).shape[0] return keyframe_idxs, num_keyframes -class LTXVAddGuide: +class LTXVAddGuide(io.ComfyNode): + NUM_PREFIX_FRAMES = 2 + PATCHIFIER = SymmetricPatchifier(1) + @classmethod - def INPUT_TYPES(s): - return {"required": {"positive": ("CONDITIONING", ), - "negative": ("CONDITIONING", ), - "vae": ("VAE",), - "latent": ("LATENT",), - "image": ("IMAGE", {"tooltip": "Image or video to condition the latent video on. Must be 8*n + 1 frames." - "If the video is not 8*n + 1 frames, it will be cropped to the nearest 8*n + 1 frames."}), - "frame_idx": ("INT", {"default": 0, "min": -9999, "max": 9999, - "tooltip": "Frame index to start the conditioning at. For single-frame images or " - "videos with 1-8 frames, any frame_idx value is acceptable. For videos with 9+ " - "frames, frame_idx must be divisible by 8, otherwise it will be rounded down to " - "the nearest multiple of 8. Negative values are counted from the end of the video."}), - "strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), - } - } + def define_schema(cls): + return io.Schema( + node_id="LTXVAddGuide", + category="conditioning/video_models", + inputs=[ + io.Conditioning.Input("positive"), + io.Conditioning.Input("negative"), + io.Vae.Input("vae"), + io.Latent.Input("latent"), + io.Image.Input( + "image", + tooltip="Image or video to condition the latent video on. Must be 8*n + 1 frames. " + "If the video is not 8*n + 1 frames, it will be cropped to the nearest 8*n + 1 frames.", + ), + io.Int.Input( + "frame_idx", + default=0, + min=-9999, + max=9999, + tooltip="Frame index to start the conditioning at. " + "For single-frame images or videos with 1-8 frames, any frame_idx value is acceptable. " + "For videos with 9+ frames, frame_idx must be divisible by 8, otherwise it will be rounded " + "down to the nearest multiple of 8. Negative values are counted from the end of the video.", + ), + io.Float.Input("strength", default=1.0, min=0.0, max=1.0, step=0.01), + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + io.Latent.Output(display_name="latent"), + ], + ) - RETURN_TYPES = ("CONDITIONING", "CONDITIONING", "LATENT") - RETURN_NAMES = ("positive", "negative", "latent") - - CATEGORY = "conditioning/video_models" - FUNCTION = "generate" - - def __init__(self): - self._num_prefix_frames = 2 - self._patchifier = SymmetricPatchifier(1) - - def encode(self, vae, latent_width, latent_height, images, scale_factors): + @classmethod + def encode(cls, vae, latent_width, latent_height, images, scale_factors): time_scale_factor, width_scale_factor, height_scale_factor = scale_factors images = images[:(images.shape[0] - 1) // time_scale_factor * time_scale_factor + 1] pixels = comfy.utils.common_upscale(images.movedim(-1, 1), latent_width * width_scale_factor, latent_height * height_scale_factor, "bilinear", crop="disabled").movedim(1, -1) @@ -129,7 +153,8 @@ class LTXVAddGuide: t = vae.encode(encode_pixels) return encode_pixels, t - def get_latent_index(self, cond, latent_length, guide_length, frame_idx, scale_factors): + @classmethod + def get_latent_index(cls, cond, latent_length, guide_length, frame_idx, scale_factors): time_scale_factor, _, _ = scale_factors _, num_keyframes = get_keyframe_idxs(cond) latent_count = latent_length - num_keyframes @@ -141,9 +166,10 @@ class LTXVAddGuide: return frame_idx, latent_idx - def add_keyframe_index(self, cond, frame_idx, guiding_latent, scale_factors): + @classmethod + def add_keyframe_index(cls, cond, frame_idx, guiding_latent, scale_factors): keyframe_idxs, _ = get_keyframe_idxs(cond) - _, latent_coords = self._patchifier.patchify(guiding_latent) + _, latent_coords = cls.PATCHIFIER.patchify(guiding_latent) pixel_coords = latent_to_pixel_coords(latent_coords, scale_factors, causal_fix=frame_idx == 0) # we need the causal fix only if we're placing the new latents at index 0 pixel_coords[:, 0] += frame_idx if keyframe_idxs is None: @@ -152,8 +178,9 @@ class LTXVAddGuide: keyframe_idxs = torch.cat([keyframe_idxs, pixel_coords], dim=2) return node_helpers.conditioning_set_values(cond, {"keyframe_idxs": keyframe_idxs}) - def append_keyframe(self, positive, negative, frame_idx, latent_image, noise_mask, guiding_latent, strength, scale_factors): - _, latent_idx = self.get_latent_index( + @classmethod + def append_keyframe(cls, positive, negative, frame_idx, latent_image, noise_mask, guiding_latent, strength, scale_factors): + _, latent_idx = cls.get_latent_index( cond=positive, latent_length=latent_image.shape[2], guide_length=guiding_latent.shape[2], @@ -162,8 +189,8 @@ class LTXVAddGuide: ) noise_mask[:, :, latent_idx:latent_idx + guiding_latent.shape[2]] = 1.0 - positive = self.add_keyframe_index(positive, frame_idx, guiding_latent, scale_factors) - negative = self.add_keyframe_index(negative, frame_idx, guiding_latent, scale_factors) + positive = cls.add_keyframe_index(positive, frame_idx, guiding_latent, scale_factors) + negative = cls.add_keyframe_index(negative, frame_idx, guiding_latent, scale_factors) mask = torch.full( (noise_mask.shape[0], 1, guiding_latent.shape[2], noise_mask.shape[3], noise_mask.shape[4]), @@ -176,7 +203,8 @@ class LTXVAddGuide: noise_mask = torch.cat([noise_mask, mask], dim=2) return positive, negative, latent_image, noise_mask - def replace_latent_frames(self, latent_image, noise_mask, guiding_latent, latent_idx, strength): + @classmethod + def replace_latent_frames(cls, latent_image, noise_mask, guiding_latent, latent_idx, strength): cond_length = guiding_latent.shape[2] assert latent_image.shape[2] >= latent_idx + cond_length, "Conditioning frames exceed the length of the latent sequence." @@ -195,20 +223,21 @@ class LTXVAddGuide: return latent_image, noise_mask - def generate(self, positive, negative, vae, latent, image, frame_idx, strength): + @classmethod + def execute(cls, positive, negative, vae, latent, image, frame_idx, strength) -> io.NodeOutput: scale_factors = vae.downscale_index_formula latent_image = latent["samples"] noise_mask = get_noise_mask(latent) _, _, latent_length, latent_height, latent_width = latent_image.shape - image, t = self.encode(vae, latent_width, latent_height, image, scale_factors) + image, t = cls.encode(vae, latent_width, latent_height, image, scale_factors) - frame_idx, latent_idx = self.get_latent_index(positive, latent_length, len(image), frame_idx, scale_factors) + frame_idx, latent_idx = cls.get_latent_index(positive, latent_length, len(image), frame_idx, scale_factors) assert latent_idx + t.shape[2] <= latent_length, "Conditioning frames exceed the length of the latent sequence." - num_prefix_frames = min(self._num_prefix_frames, t.shape[2]) + num_prefix_frames = min(cls.NUM_PREFIX_FRAMES, t.shape[2]) - positive, negative, latent_image, noise_mask = self.append_keyframe( + positive, negative, latent_image, noise_mask = cls.append_keyframe( positive, negative, frame_idx, @@ -223,9 +252,9 @@ class LTXVAddGuide: t = t[:, :, num_prefix_frames:] if t.shape[2] == 0: - return (positive, negative, {"samples": latent_image, "noise_mask": noise_mask},) + return io.NodeOutput(positive, negative, {"samples": latent_image, "noise_mask": noise_mask}) - latent_image, noise_mask = self.replace_latent_frames( + latent_image, noise_mask = cls.replace_latent_frames( latent_image, noise_mask, t, @@ -233,34 +262,35 @@ class LTXVAddGuide: strength, ) - return (positive, negative, {"samples": latent_image, "noise_mask": noise_mask},) + return io.NodeOutput(positive, negative, {"samples": latent_image, "noise_mask": noise_mask}) -class LTXVCropGuides: +class LTXVCropGuides(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"positive": ("CONDITIONING", ), - "negative": ("CONDITIONING", ), - "latent": ("LATENT",), - } - } + def define_schema(cls): + return io.Schema( + node_id="LTXVCropGuides", + category="conditioning/video_models", + inputs=[ + io.Conditioning.Input("positive"), + io.Conditioning.Input("negative"), + io.Latent.Input("latent"), + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + io.Latent.Output(display_name="latent"), + ], + ) - RETURN_TYPES = ("CONDITIONING", "CONDITIONING", "LATENT") - RETURN_NAMES = ("positive", "negative", "latent") - - CATEGORY = "conditioning/video_models" - FUNCTION = "crop" - - def __init__(self): - self._patchifier = SymmetricPatchifier(1) - - def crop(self, positive, negative, latent): + @classmethod + def execute(cls, positive, negative, latent) -> io.NodeOutput: latent_image = latent["samples"].clone() noise_mask = get_noise_mask(latent) _, num_keyframes = get_keyframe_idxs(positive) if num_keyframes == 0: - return (positive, negative, {"samples": latent_image, "noise_mask": noise_mask},) + return io.NodeOutput(positive, negative, {"samples": latent_image, "noise_mask": noise_mask},) latent_image = latent_image[:, :, :-num_keyframes] noise_mask = noise_mask[:, :, :-num_keyframes] @@ -268,44 +298,52 @@ class LTXVCropGuides: positive = node_helpers.conditioning_set_values(positive, {"keyframe_idxs": None}) negative = node_helpers.conditioning_set_values(negative, {"keyframe_idxs": None}) - return (positive, negative, {"samples": latent_image, "noise_mask": noise_mask},) + return io.NodeOutput(positive, negative, {"samples": latent_image, "noise_mask": noise_mask}) -class LTXVConditioning: +class LTXVConditioning(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"positive": ("CONDITIONING", ), - "negative": ("CONDITIONING", ), - "frame_rate": ("FLOAT", {"default": 25.0, "min": 0.0, "max": 1000.0, "step": 0.01}), - }} - RETURN_TYPES = ("CONDITIONING", "CONDITIONING") - RETURN_NAMES = ("positive", "negative") - FUNCTION = "append" + def define_schema(cls): + return io.Schema( + node_id="LTXVConditioning", + category="conditioning/video_models", + inputs=[ + io.Conditioning.Input("positive"), + io.Conditioning.Input("negative"), + io.Float.Input("frame_rate", default=25.0, min=0.0, max=1000.0, step=0.01), + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + ], + ) - CATEGORY = "conditioning/video_models" - - def append(self, positive, negative, frame_rate): + @classmethod + def execute(cls, positive, negative, frame_rate) -> io.NodeOutput: positive = node_helpers.conditioning_set_values(positive, {"frame_rate": frame_rate}) negative = node_helpers.conditioning_set_values(negative, {"frame_rate": frame_rate}) - return (positive, negative) + return io.NodeOutput(positive, negative) -class ModelSamplingLTXV: +class ModelSamplingLTXV(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "model": ("MODEL",), - "max_shift": ("FLOAT", {"default": 2.05, "min": 0.0, "max": 100.0, "step":0.01}), - "base_shift": ("FLOAT", {"default": 0.95, "min": 0.0, "max": 100.0, "step":0.01}), - }, - "optional": {"latent": ("LATENT",), } - } + def define_schema(cls): + return io.Schema( + node_id="ModelSamplingLTXV", + category="advanced/model", + inputs=[ + io.Model.Input("model"), + io.Float.Input("max_shift", default=2.05, min=0.0, max=100.0, step=0.01), + io.Float.Input("base_shift", default=0.95, min=0.0, max=100.0, step=0.01), + io.Latent.Input("latent", optional=True), + ], + outputs=[ + io.Model.Output(), + ], + ) - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" - - CATEGORY = "advanced/model" - - def patch(self, model, max_shift, base_shift, latent=None): + @classmethod + def execute(cls, model, max_shift, base_shift, latent=None) -> io.NodeOutput: m = model.clone() if latent is None: @@ -329,37 +367,41 @@ class ModelSamplingLTXV: model_sampling.set_parameters(shift=shift) m.add_object_patch("model_sampling", model_sampling) - return (m, ) + return io.NodeOutput(m) -class LTXVScheduler: +class LTXVScheduler(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": - {"steps": ("INT", {"default": 20, "min": 1, "max": 10000}), - "max_shift": ("FLOAT", {"default": 2.05, "min": 0.0, "max": 100.0, "step":0.01}), - "base_shift": ("FLOAT", {"default": 0.95, "min": 0.0, "max": 100.0, "step":0.01}), - "stretch": ("BOOLEAN", { - "default": True, - "tooltip": "Stretch the sigmas to be in the range [terminal, 1]." - }), - "terminal": ( - "FLOAT", - { - "default": 0.1, "min": 0.0, "max": 0.99, "step": 0.01, - "tooltip": "The terminal value of the sigmas after stretching." - }, - ), - }, - "optional": {"latent": ("LATENT",), } - } + def define_schema(cls): + return io.Schema( + node_id="LTXVScheduler", + category="sampling/custom_sampling/schedulers", + inputs=[ + io.Int.Input("steps", default=20, min=1, max=10000), + io.Float.Input("max_shift", default=2.05, min=0.0, max=100.0, step=0.01), + io.Float.Input("base_shift", default=0.95, min=0.0, max=100.0, step=0.01), + io.Boolean.Input( + id="stretch", + default=True, + tooltip="Stretch the sigmas to be in the range [terminal, 1].", + ), + io.Float.Input( + id="terminal", + default=0.1, + min=0.0, + max=0.99, + step=0.01, + tooltip="The terminal value of the sigmas after stretching.", + ), + io.Latent.Input("latent", optional=True), + ], + outputs=[ + io.Sigmas.Output(), + ], + ) - RETURN_TYPES = ("SIGMAS",) - CATEGORY = "sampling/custom_sampling/schedulers" - - FUNCTION = "get_sigmas" - - def get_sigmas(self, steps, max_shift, base_shift, stretch, terminal, latent=None): + @classmethod + def execute(cls, steps, max_shift, base_shift, stretch, terminal, latent=None) -> io.NodeOutput: if latent is None: tokens = 4096 else: @@ -389,7 +431,7 @@ class LTXVScheduler: stretched = 1.0 - (one_minus_z / scale_factor) sigmas[non_zero_mask] = stretched - return (sigmas,) + return io.NodeOutput(sigmas) def encode_single_frame(output_file, image_array: np.ndarray, crf): container = av.open(output_file, "w", format="mp4") @@ -423,52 +465,54 @@ def preprocess(image: torch.Tensor, crf=29): return image image_array = (image[:(image.shape[0] // 2) * 2, :(image.shape[1] // 2) * 2] * 255.0).byte().cpu().numpy() - with io.BytesIO() as output_file: + with BytesIO() as output_file: encode_single_frame(output_file, image_array, crf) video_bytes = output_file.getvalue() - with io.BytesIO(video_bytes) as video_file: + with BytesIO(video_bytes) as video_file: image_array = decode_single_frame(video_file) tensor = torch.tensor(image_array, dtype=image.dtype, device=image.device) / 255.0 return tensor -class LTXVPreprocess: +class LTXVPreprocess(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": ("IMAGE",), - "img_compression": ( - "INT", - { - "default": 35, - "min": 0, - "max": 100, - "tooltip": "Amount of compression to apply on image.", - }, + def define_schema(cls): + return io.Schema( + node_id="LTXVPreprocess", + category="image", + inputs=[ + io.Image.Input("image"), + io.Int.Input( + id="img_compression", default=35, min=0, max=100, tooltip="Amount of compression to apply on image." ), - } - } + ], + outputs=[ + io.Image.Output(display_name="output_image"), + ], + ) - FUNCTION = "preprocess" - RETURN_TYPES = ("IMAGE",) - RETURN_NAMES = ("output_image",) - CATEGORY = "image" - - def preprocess(self, image, img_compression): + @classmethod + def execute(cls, image, img_compression) -> io.NodeOutput: output_images = [] for i in range(image.shape[0]): output_images.append(preprocess(image[i], img_compression)) - return (torch.stack(output_images),) + return io.NodeOutput(torch.stack(output_images)) -NODE_CLASS_MAPPINGS = { - "EmptyLTXVLatentVideo": EmptyLTXVLatentVideo, - "LTXVImgToVideo": LTXVImgToVideo, - "ModelSamplingLTXV": ModelSamplingLTXV, - "LTXVConditioning": LTXVConditioning, - "LTXVScheduler": LTXVScheduler, - "LTXVAddGuide": LTXVAddGuide, - "LTXVPreprocess": LTXVPreprocess, - "LTXVCropGuides": LTXVCropGuides, -} +class LtxvExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + EmptyLTXVLatentVideo, + LTXVImgToVideo, + ModelSamplingLTXV, + LTXVConditioning, + LTXVScheduler, + LTXVAddGuide, + LTXVPreprocess, + LTXVCropGuides, + ] + + +async def comfy_entrypoint() -> LtxvExtension: + return LtxvExtension() From e4f99b479a19730bea890567129f4032b4dd4787 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 1 Oct 2025 22:20:30 +0300 Subject: [PATCH 073/114] convert nodes_ip2p.pt to V3 schema (#10097) --- comfy_extras/nodes_ip2p.py | 54 +++++++++++++++++++++++++------------- 1 file changed, 36 insertions(+), 18 deletions(-) diff --git a/comfy_extras/nodes_ip2p.py b/comfy_extras/nodes_ip2p.py index c2e70a84c..78f29915d 100644 --- a/comfy_extras/nodes_ip2p.py +++ b/comfy_extras/nodes_ip2p.py @@ -1,21 +1,30 @@ import torch -class InstructPixToPixConditioning: +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io + + +class InstructPixToPixConditioning(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"positive": ("CONDITIONING", ), - "negative": ("CONDITIONING", ), - "vae": ("VAE", ), - "pixels": ("IMAGE", ), - }} + def define_schema(cls): + return io.Schema( + node_id="InstructPixToPixConditioning", + category="conditioning/instructpix2pix", + inputs=[ + io.Conditioning.Input("positive"), + io.Conditioning.Input("negative"), + io.Vae.Input("vae"), + io.Image.Input("pixels"), + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + io.Latent.Output(display_name="latent"), + ], + ) - RETURN_TYPES = ("CONDITIONING","CONDITIONING","LATENT") - RETURN_NAMES = ("positive", "negative", "latent") - FUNCTION = "encode" - - CATEGORY = "conditioning/instructpix2pix" - - def encode(self, positive, negative, pixels, vae): + @classmethod + def execute(cls, positive, negative, pixels, vae) -> io.NodeOutput: x = (pixels.shape[1] // 8) * 8 y = (pixels.shape[2] // 8) * 8 @@ -38,8 +47,17 @@ class InstructPixToPixConditioning: n = [t[0], d] c.append(n) out.append(c) - return (out[0], out[1], out_latent) + return io.NodeOutput(out[0], out[1], out_latent) + + +class InstructPix2PixExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + InstructPixToPixConditioning, + ] + + +async def comfy_entrypoint() -> InstructPix2PixExtension: + return InstructPix2PixExtension() -NODE_CLASS_MAPPINGS = { - "InstructPixToPixConditioning": InstructPixToPixConditioning, -} From a6f83a4a1a70d720c16d66feb5d87fee5998acdf Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Wed, 1 Oct 2025 14:19:13 -0700 Subject: [PATCH 074/114] Support the new hunyuan vae. (#10150) --- comfy/ldm/hunyuan_video/vae_refiner.py | 112 ++++++++++++++++--------- comfy/sd.py | 70 ++++++++++------ 2 files changed, 116 insertions(+), 66 deletions(-) diff --git a/comfy/ldm/hunyuan_video/vae_refiner.py b/comfy/ldm/hunyuan_video/vae_refiner.py index c6f742710..c2a0b507d 100644 --- a/comfy/ldm/hunyuan_video/vae_refiner.py +++ b/comfy/ldm/hunyuan_video/vae_refiner.py @@ -1,7 +1,7 @@ import torch import torch.nn as nn import torch.nn.functional as F -from comfy.ldm.modules.diffusionmodules.model import ResnetBlock, AttnBlock, VideoConv3d +from comfy.ldm.modules.diffusionmodules.model import ResnetBlock, AttnBlock, VideoConv3d, Normalize import comfy.ops import comfy.ldm.models.autoencoder ops = comfy.ops.disable_weight_init @@ -17,11 +17,12 @@ class RMS_norm(nn.Module): return F.normalize(x, dim=1) * self.scale * self.gamma class DnSmpl(nn.Module): - def __init__(self, ic, oc, tds=True): + def __init__(self, ic, oc, tds=True, refiner_vae=True, op=VideoConv3d): super().__init__() fct = 2 * 2 * 2 if tds else 1 * 2 * 2 assert oc % fct == 0 - self.conv = VideoConv3d(ic, oc // fct, kernel_size=3) + self.conv = op(ic, oc // fct, kernel_size=3, stride=1, padding=1) + self.refiner_vae = refiner_vae self.tds = tds self.gs = fct * ic // oc @@ -30,7 +31,7 @@ class DnSmpl(nn.Module): r1 = 2 if self.tds else 1 h = self.conv(x) - if self.tds: + if self.tds and self.refiner_vae: hf = h[:, :, :1, :, :] b, c, f, ht, wd = hf.shape hf = hf.reshape(b, c, f, ht // 2, 2, wd // 2, 2) @@ -66,6 +67,7 @@ class DnSmpl(nn.Module): sc = torch.cat([xf, xn], dim=2) else: b, c, frms, ht, wd = h.shape + nf = frms // r1 h = h.reshape(b, c, nf, r1, ht // 2, 2, wd // 2, 2) h = h.permute(0, 3, 5, 7, 1, 2, 4, 6) @@ -83,10 +85,11 @@ class DnSmpl(nn.Module): class UpSmpl(nn.Module): - def __init__(self, ic, oc, tus=True): + def __init__(self, ic, oc, tus=True, refiner_vae=True, op=VideoConv3d): super().__init__() fct = 2 * 2 * 2 if tus else 1 * 2 * 2 - self.conv = VideoConv3d(ic, oc * fct, kernel_size=3) + self.conv = op(ic, oc * fct, kernel_size=3, stride=1, padding=1) + self.refiner_vae = refiner_vae self.tus = tus self.rp = fct * oc // ic @@ -95,7 +98,7 @@ class UpSmpl(nn.Module): r1 = 2 if self.tus else 1 h = self.conv(x) - if self.tus: + if self.tus and self.refiner_vae: hf = h[:, :, :1, :, :] b, c, f, ht, wd = hf.shape nc = c // (2 * 2) @@ -148,43 +151,56 @@ class UpSmpl(nn.Module): class Encoder(nn.Module): def __init__(self, in_channels, z_channels, block_out_channels, num_res_blocks, - ffactor_spatial, ffactor_temporal, downsample_match_channel=True, **_): + ffactor_spatial, ffactor_temporal, downsample_match_channel=True, refiner_vae=True, **_): super().__init__() self.z_channels = z_channels self.block_out_channels = block_out_channels self.num_res_blocks = num_res_blocks - self.conv_in = VideoConv3d(in_channels, block_out_channels[0], 3, 1, 1) + self.ffactor_temporal = ffactor_temporal + + self.refiner_vae = refiner_vae + if self.refiner_vae: + conv_op = VideoConv3d + norm_op = RMS_norm + else: + conv_op = ops.Conv3d + norm_op = Normalize + + self.conv_in = conv_op(in_channels, block_out_channels[0], 3, 1, 1) self.down = nn.ModuleList() ch = block_out_channels[0] depth = (ffactor_spatial >> 1).bit_length() - depth_temporal = ((ffactor_spatial // ffactor_temporal) >> 1).bit_length() + depth_temporal = ((ffactor_spatial // self.ffactor_temporal) >> 1).bit_length() for i, tgt in enumerate(block_out_channels): stage = nn.Module() stage.block = nn.ModuleList([ResnetBlock(in_channels=ch if j == 0 else tgt, out_channels=tgt, temb_channels=0, - conv_op=VideoConv3d, norm_op=RMS_norm) + conv_op=conv_op, norm_op=norm_op) for j in range(num_res_blocks)]) ch = tgt if i < depth: nxt = block_out_channels[i + 1] if i + 1 < len(block_out_channels) and downsample_match_channel else ch - stage.downsample = DnSmpl(ch, nxt, tds=i >= depth_temporal) + stage.downsample = DnSmpl(ch, nxt, tds=i >= depth_temporal, refiner_vae=self.refiner_vae, op=conv_op) ch = nxt self.down.append(stage) self.mid = nn.Module() - self.mid.block_1 = ResnetBlock(in_channels=ch, out_channels=ch, temb_channels=0, conv_op=VideoConv3d, norm_op=RMS_norm) - self.mid.attn_1 = AttnBlock(ch, conv_op=ops.Conv3d, norm_op=RMS_norm) - self.mid.block_2 = ResnetBlock(in_channels=ch, out_channels=ch, temb_channels=0, conv_op=VideoConv3d, norm_op=RMS_norm) + self.mid.block_1 = ResnetBlock(in_channels=ch, out_channels=ch, temb_channels=0, conv_op=conv_op, norm_op=norm_op) + self.mid.attn_1 = AttnBlock(ch, conv_op=ops.Conv3d, norm_op=norm_op) + self.mid.block_2 = ResnetBlock(in_channels=ch, out_channels=ch, temb_channels=0, conv_op=conv_op, norm_op=norm_op) - self.norm_out = RMS_norm(ch) - self.conv_out = VideoConv3d(ch, z_channels << 1, 3, 1, 1) + self.norm_out = norm_op(ch) + self.conv_out = conv_op(ch, z_channels << 1, 3, 1, 1) self.regul = comfy.ldm.models.autoencoder.DiagonalGaussianRegularizer() def forward(self, x): + if not self.refiner_vae and x.shape[2] == 1: + x = x.expand(-1, -1, self.ffactor_temporal, -1, -1) + x = self.conv_in(x) for stage in self.down: @@ -200,31 +216,42 @@ class Encoder(nn.Module): skip = x.view(b, c // grp, grp, t, h, w).mean(2) out = self.conv_out(F.silu(self.norm_out(x))) + skip - out = self.regul(out)[0] - out = torch.cat((out[:, :, :1], out), dim=2) - out = out.permute(0, 2, 1, 3, 4) - b, f_times_2, c, h, w = out.shape - out = out.reshape(b, f_times_2 // 2, 2 * c, h, w) - out = out.permute(0, 2, 1, 3, 4).contiguous() + if self.refiner_vae: + out = self.regul(out)[0] + + out = torch.cat((out[:, :, :1], out), dim=2) + out = out.permute(0, 2, 1, 3, 4) + b, f_times_2, c, h, w = out.shape + out = out.reshape(b, f_times_2 // 2, 2 * c, h, w) + out = out.permute(0, 2, 1, 3, 4).contiguous() + return out class Decoder(nn.Module): def __init__(self, z_channels, out_channels, block_out_channels, num_res_blocks, - ffactor_spatial, ffactor_temporal, upsample_match_channel=True, **_): + ffactor_spatial, ffactor_temporal, upsample_match_channel=True, refiner_vae=True, **_): super().__init__() block_out_channels = block_out_channels[::-1] self.z_channels = z_channels self.block_out_channels = block_out_channels self.num_res_blocks = num_res_blocks + self.refiner_vae = refiner_vae + if self.refiner_vae: + conv_op = VideoConv3d + norm_op = RMS_norm + else: + conv_op = ops.Conv3d + norm_op = Normalize + ch = block_out_channels[0] - self.conv_in = VideoConv3d(z_channels, ch, 3) + self.conv_in = conv_op(z_channels, ch, kernel_size=3, stride=1, padding=1) self.mid = nn.Module() - self.mid.block_1 = ResnetBlock(in_channels=ch, out_channels=ch, temb_channels=0, conv_op=VideoConv3d, norm_op=RMS_norm) - self.mid.attn_1 = AttnBlock(ch, conv_op=ops.Conv3d, norm_op=RMS_norm) - self.mid.block_2 = ResnetBlock(in_channels=ch, out_channels=ch, temb_channels=0, conv_op=VideoConv3d, norm_op=RMS_norm) + self.mid.block_1 = ResnetBlock(in_channels=ch, out_channels=ch, temb_channels=0, conv_op=conv_op, norm_op=norm_op) + self.mid.attn_1 = AttnBlock(ch, conv_op=ops.Conv3d, norm_op=norm_op) + self.mid.block_2 = ResnetBlock(in_channels=ch, out_channels=ch, temb_channels=0, conv_op=conv_op, norm_op=norm_op) self.up = nn.ModuleList() depth = (ffactor_spatial >> 1).bit_length() @@ -235,25 +262,26 @@ class Decoder(nn.Module): stage.block = nn.ModuleList([ResnetBlock(in_channels=ch if j == 0 else tgt, out_channels=tgt, temb_channels=0, - conv_op=VideoConv3d, norm_op=RMS_norm) + conv_op=conv_op, norm_op=norm_op) for j in range(num_res_blocks + 1)]) ch = tgt if i < depth: nxt = block_out_channels[i + 1] if i + 1 < len(block_out_channels) and upsample_match_channel else ch - stage.upsample = UpSmpl(ch, nxt, tus=i < depth_temporal) + stage.upsample = UpSmpl(ch, nxt, tus=i < depth_temporal, refiner_vae=self.refiner_vae, op=conv_op) ch = nxt self.up.append(stage) - self.norm_out = RMS_norm(ch) - self.conv_out = VideoConv3d(ch, out_channels, 3) + self.norm_out = norm_op(ch) + self.conv_out = conv_op(ch, out_channels, 3, stride=1, padding=1) def forward(self, z): - z = z.permute(0, 2, 1, 3, 4) - b, f, c, h, w = z.shape - z = z.reshape(b, f, 2, c // 2, h, w) - z = z.permute(0, 1, 2, 3, 4, 5).reshape(b, f * 2, c // 2, h, w) - z = z.permute(0, 2, 1, 3, 4) - z = z[:, :, 1:] + if self.refiner_vae: + z = z.permute(0, 2, 1, 3, 4) + b, f, c, h, w = z.shape + z = z.reshape(b, f, 2, c // 2, h, w) + z = z.permute(0, 1, 2, 3, 4, 5).reshape(b, f * 2, c // 2, h, w) + z = z.permute(0, 2, 1, 3, 4) + z = z[:, :, 1:] x = self.conv_in(z) + z.repeat_interleave(self.block_out_channels[0] // self.z_channels, 1) x = self.mid.block_2(self.mid.attn_1(self.mid.block_1(x))) @@ -264,4 +292,10 @@ class Decoder(nn.Module): if hasattr(stage, 'upsample'): x = stage.upsample(x) - return self.conv_out(F.silu(self.norm_out(x))) + out = self.conv_out(F.silu(self.norm_out(x))) + + if not self.refiner_vae: + if z.shape[-3] == 1: + out = out[:, :, -1:] + + return out diff --git a/comfy/sd.py b/comfy/sd.py index 2df340739..873ad20f2 100644 --- a/comfy/sd.py +++ b/comfy/sd.py @@ -332,35 +332,51 @@ class VAE: self.first_stage_model = StageC_coder() self.downscale_ratio = 32 self.latent_channels = 16 - elif "decoder.conv_in.weight" in sd and sd['decoder.conv_in.weight'].shape[1] == 64: - ddconfig = {"block_out_channels": [128, 256, 512, 512, 1024, 1024], "in_channels": 3, "out_channels": 3, "num_res_blocks": 2, "ffactor_spatial": 32, "downsample_match_channel": True, "upsample_match_channel": True} - self.latent_channels = ddconfig['z_channels'] = sd["decoder.conv_in.weight"].shape[1] - self.downscale_ratio = 32 - self.upscale_ratio = 32 - self.working_dtypes = [torch.float16, torch.bfloat16, torch.float32] - self.first_stage_model = AutoencodingEngine(regularizer_config={'target': "comfy.ldm.models.autoencoder.DiagonalGaussianRegularizer"}, - encoder_config={'target': "comfy.ldm.hunyuan_video.vae.Encoder", 'params': ddconfig}, - decoder_config={'target': "comfy.ldm.hunyuan_video.vae.Decoder", 'params': ddconfig}) - - self.memory_used_encode = lambda shape, dtype: (700 * shape[2] * shape[3]) * model_management.dtype_size(dtype) - self.memory_used_decode = lambda shape, dtype: (700 * shape[2] * shape[3] * 32 * 32) * model_management.dtype_size(dtype) - elif "decoder.conv_in.weight" in sd: - #default SD1.x/SD2.x VAE parameters - ddconfig = {'double_z': True, 'z_channels': 4, 'resolution': 256, 'in_channels': 3, 'out_ch': 3, 'ch': 128, 'ch_mult': [1, 2, 4, 4], 'num_res_blocks': 2, 'attn_resolutions': [], 'dropout': 0.0} - - if 'encoder.down.2.downsample.conv.weight' not in sd and 'decoder.up.3.upsample.conv.weight' not in sd: #Stable diffusion x4 upscaler VAE - ddconfig['ch_mult'] = [1, 2, 4] - self.downscale_ratio = 4 - self.upscale_ratio = 4 - - self.latent_channels = ddconfig['z_channels'] = sd["decoder.conv_in.weight"].shape[1] - if 'post_quant_conv.weight' in sd: - self.first_stage_model = AutoencoderKL(ddconfig=ddconfig, embed_dim=sd['post_quant_conv.weight'].shape[1]) - else: + if sd['decoder.conv_in.weight'].shape[1] == 64: + ddconfig = {"block_out_channels": [128, 256, 512, 512, 1024, 1024], "in_channels": 3, "out_channels": 3, "num_res_blocks": 2, "ffactor_spatial": 32, "downsample_match_channel": True, "upsample_match_channel": True} + self.latent_channels = ddconfig['z_channels'] = sd["decoder.conv_in.weight"].shape[1] + self.downscale_ratio = 32 + self.upscale_ratio = 32 + self.working_dtypes = [torch.float16, torch.bfloat16, torch.float32] self.first_stage_model = AutoencodingEngine(regularizer_config={'target': "comfy.ldm.models.autoencoder.DiagonalGaussianRegularizer"}, - encoder_config={'target': "comfy.ldm.modules.diffusionmodules.model.Encoder", 'params': ddconfig}, - decoder_config={'target': "comfy.ldm.modules.diffusionmodules.model.Decoder", 'params': ddconfig}) + encoder_config={'target': "comfy.ldm.hunyuan_video.vae.Encoder", 'params': ddconfig}, + decoder_config={'target': "comfy.ldm.hunyuan_video.vae.Decoder", 'params': ddconfig}) + + self.memory_used_encode = lambda shape, dtype: (700 * shape[2] * shape[3]) * model_management.dtype_size(dtype) + self.memory_used_decode = lambda shape, dtype: (700 * shape[2] * shape[3] * 32 * 32) * model_management.dtype_size(dtype) + elif sd['decoder.conv_in.weight'].shape[1] == 32: + ddconfig = {"block_out_channels": [128, 256, 512, 1024, 1024], "in_channels": 3, "out_channels": 3, "num_res_blocks": 2, "ffactor_spatial": 16, "ffactor_temporal": 4, "downsample_match_channel": True, "upsample_match_channel": True, "refiner_vae": False} + self.latent_channels = ddconfig['z_channels'] = sd["decoder.conv_in.weight"].shape[1] + self.working_dtypes = [torch.float16, torch.bfloat16, torch.float32] + self.upscale_ratio = (lambda a: max(0, a * 4 - 3), 16, 16) + self.upscale_index_formula = (4, 16, 16) + self.downscale_ratio = (lambda a: max(0, math.floor((a + 3) / 4)), 16, 16) + self.downscale_index_formula = (4, 16, 16) + self.latent_dim = 3 + self.not_video = True + self.first_stage_model = AutoencodingEngine(regularizer_config={'target': "comfy.ldm.models.autoencoder.DiagonalGaussianRegularizer"}, + encoder_config={'target': "comfy.ldm.hunyuan_video.vae_refiner.Encoder", 'params': ddconfig}, + decoder_config={'target': "comfy.ldm.hunyuan_video.vae_refiner.Decoder", 'params': ddconfig}) + + self.memory_used_encode = lambda shape, dtype: (2800 * shape[-2] * shape[-1]) * model_management.dtype_size(dtype) + self.memory_used_decode = lambda shape, dtype: (2800 * shape[-3] * shape[-2] * shape[-1] * 16 * 16) * model_management.dtype_size(dtype) + else: + #default SD1.x/SD2.x VAE parameters + ddconfig = {'double_z': True, 'z_channels': 4, 'resolution': 256, 'in_channels': 3, 'out_ch': 3, 'ch': 128, 'ch_mult': [1, 2, 4, 4], 'num_res_blocks': 2, 'attn_resolutions': [], 'dropout': 0.0} + + if 'encoder.down.2.downsample.conv.weight' not in sd and 'decoder.up.3.upsample.conv.weight' not in sd: #Stable diffusion x4 upscaler VAE + ddconfig['ch_mult'] = [1, 2, 4] + self.downscale_ratio = 4 + self.upscale_ratio = 4 + + self.latent_channels = ddconfig['z_channels'] = sd["decoder.conv_in.weight"].shape[1] + if 'post_quant_conv.weight' in sd: + self.first_stage_model = AutoencoderKL(ddconfig=ddconfig, embed_dim=sd['post_quant_conv.weight'].shape[1]) + else: + self.first_stage_model = AutoencodingEngine(regularizer_config={'target': "comfy.ldm.models.autoencoder.DiagonalGaussianRegularizer"}, + encoder_config={'target': "comfy.ldm.modules.diffusionmodules.model.Encoder", 'params': ddconfig}, + decoder_config={'target': "comfy.ldm.modules.diffusionmodules.model.Decoder", 'params': ddconfig}) elif "decoder.layers.1.layers.0.beta" in sd: self.first_stage_model = AudioOobleckVAE() self.memory_used_encode = lambda shape, dtype: (1000 * shape[2]) * model_management.dtype_size(dtype) From bb32d4ec3141333df26fcdaee0c3c08e41b7b249 Mon Sep 17 00:00:00 2001 From: Koratahiu Date: Thu, 2 Oct 2025 00:59:07 +0300 Subject: [PATCH 075/114] feat: Add Epsilon Scaling node for exposure bias correction (#10132) --- comfy_extras/nodes_eps.py | 60 +++++++++++++++++++++++++++++++++++++++ nodes.py | 1 + 2 files changed, 61 insertions(+) create mode 100644 comfy_extras/nodes_eps.py diff --git a/comfy_extras/nodes_eps.py b/comfy_extras/nodes_eps.py new file mode 100644 index 000000000..c8818f096 --- /dev/null +++ b/comfy_extras/nodes_eps.py @@ -0,0 +1,60 @@ +class EpsilonScaling: + """ + Implements the Epsilon Scaling method from 'Elucidating the Exposure Bias in Diffusion Models' + (https://arxiv.org/abs/2308.15321v6). + + This method mitigates exposure bias by scaling the predicted noise during sampling, + which can significantly improve sample quality. This implementation uses the "uniform schedule" + recommended by the paper for its practicality and effectiveness. + """ + @classmethod + def INPUT_TYPES(s): + return { + "required": { + "model": ("MODEL",), + "scaling_factor": ("FLOAT", { + "default": 1.005, + "min": 0.5, + "max": 1.5, + "step": 0.001, + "display": "number" + }), + } + } + + RETURN_TYPES = ("MODEL",) + FUNCTION = "patch" + + CATEGORY = "model_patches/unet" + + def patch(self, model, scaling_factor): + # Prevent division by zero, though the UI's min value should prevent this. + if scaling_factor == 0: + scaling_factor = 1e-9 + + def epsilon_scaling_function(args): + """ + This function is applied after the CFG guidance has been calculated. + It recalculates the denoised latent by scaling the predicted noise. + """ + denoised = args["denoised"] + x = args["input"] + + noise_pred = x - denoised + + scaled_noise_pred = noise_pred / scaling_factor + + new_denoised = x - scaled_noise_pred + + return new_denoised + + # Clone the model patcher to avoid modifying the original model in place + model_clone = model.clone() + + model_clone.set_model_sampler_post_cfg_function(epsilon_scaling_function) + + return (model_clone,) + +NODE_CLASS_MAPPINGS = { + "Epsilon Scaling": EpsilonScaling +} diff --git a/nodes.py b/nodes.py index 1a6784b68..88d712993 100644 --- a/nodes.py +++ b/nodes.py @@ -2297,6 +2297,7 @@ async def init_builtin_extra_nodes(): "nodes_gits.py", "nodes_controlnet.py", "nodes_hunyuan.py", + "nodes_eps.py", "nodes_flux.py", "nodes_lora_extract.py", "nodes_torch_compile.py", From 911331c06c16aa80633c5438c58edb32dbfdff50 Mon Sep 17 00:00:00 2001 From: rattus128 <46076784+rattus128@users.noreply.github.com> Date: Thu, 2 Oct 2025 08:40:28 +1000 Subject: [PATCH 076/114] sd: fix VAE tiled fallback VRAM leak (#10139) When the VAE catches this VRAM OOM, it launches the fallback logic straight from the exception context. Python however refs the entire call stack that caused the exception including any local variables for the sake of exception report and debugging. In the case of tensors, this can hold on the references to GBs of VRAM and inhibit the VRAM allocated from freeing them. So dump the except context completely before going back to the VAE via the tiler by getting out of the except block with nothing but a flag. The greately increases the reliability of the tiler fallback, especially on low VRAM cards, as with the bug, if the leak randomly leaked more than the headroom needed for a single tile, the tiler would fallback would OOM and fail the flow. --- comfy/sd.py | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/comfy/sd.py b/comfy/sd.py index 873ad20f2..be225ad03 100644 --- a/comfy/sd.py +++ b/comfy/sd.py @@ -652,6 +652,7 @@ class VAE: def decode(self, samples_in, vae_options={}): self.throw_exception_if_invalid() pixel_samples = None + do_tile = False try: memory_used = self.memory_used_decode(samples_in.shape, self.vae_dtype) model_management.load_models_gpu([self.patcher], memory_required=memory_used, force_full_load=self.disable_offload) @@ -667,6 +668,13 @@ class VAE: pixel_samples[x:x+batch_number] = out except model_management.OOM_EXCEPTION: logging.warning("Warning: Ran out of memory when regular VAE decoding, retrying with tiled VAE decoding.") + #NOTE: We don't know what tensors were allocated to stack variables at the time of the + #exception and the exception itself refs them all until we get out of this except block. + #So we just set a flag for tiler fallback so that tensor gc can happen once the + #exception is fully off the books. + do_tile = True + + if do_tile: dims = samples_in.ndim - 2 if dims == 1 or self.extra_1d_channel is not None: pixel_samples = self.decode_tiled_1d(samples_in) @@ -713,6 +721,7 @@ class VAE: self.throw_exception_if_invalid() pixel_samples = self.vae_encode_crop_pixels(pixel_samples) pixel_samples = pixel_samples.movedim(-1, 1) + do_tile = False if self.latent_dim == 3 and pixel_samples.ndim < 5: if not self.not_video: pixel_samples = pixel_samples.movedim(1, 0).unsqueeze(0) @@ -734,6 +743,13 @@ class VAE: except model_management.OOM_EXCEPTION: logging.warning("Warning: Ran out of memory when regular VAE encoding, retrying with tiled VAE encoding.") + #NOTE: We don't know what tensors were allocated to stack variables at the time of the + #exception and the exception itself refs them all until we get out of this except block. + #So we just set a flag for tiler fallback so that tensor gc can happen once the + #exception is fully off the books. + do_tile = True + + if do_tile: if self.latent_dim == 3: tile = 256 overlap = tile // 4 From 4965c0e2acf39d84e82cb63dd6cc4400299d0a61 Mon Sep 17 00:00:00 2001 From: rattus128 <46076784+rattus128@users.noreply.github.com> Date: Thu, 2 Oct 2025 08:42:16 +1000 Subject: [PATCH 077/114] WAN: Fix cache VRAM leak on error (#10141) If this suffers an exception (such as a VRAM oom) it will leave the encode() and decode() methods which skips the cleanup of the WAN feature cache. The comfy node cache then ultimately keeps a reference this object which is in turn reffing large tensors from the failed execution. The feature cache is currently setup at a class variable on the encoder/decoder however, the encode and decode functions always clear it on both entry and exit of normal execution. Its likely the design intent is this is usable as a streaming encoder where the input comes in batches, however the functions as they are today don't support that. So simplify by bringing the cache back to local variable, so that if it does VRAM OOM the cache itself is properly garbage when the encode()/decode() functions dissappear from the stack. --- comfy/ldm/wan/vae.py | 37 ++++++++++++++----------------------- 1 file changed, 14 insertions(+), 23 deletions(-) diff --git a/comfy/ldm/wan/vae.py b/comfy/ldm/wan/vae.py index 791596938..ccbb25822 100644 --- a/comfy/ldm/wan/vae.py +++ b/comfy/ldm/wan/vae.py @@ -468,55 +468,46 @@ class WanVAE(nn.Module): attn_scales, self.temperal_upsample, dropout) def encode(self, x): - self.clear_cache() + conv_idx = [0] + feat_map = [None] * count_conv3d(self.decoder) ## cache t = x.shape[2] iter_ = 1 + (t - 1) // 4 ## 对encode输入的x，按时间拆分为1、4、4、4.... for i in range(iter_): - self._enc_conv_idx = [0] + conv_idx = [0] if i == 0: out = self.encoder( x[:, :, :1, :, :], - feat_cache=self._enc_feat_map, - feat_idx=self._enc_conv_idx) + feat_cache=feat_map, + feat_idx=conv_idx) else: out_ = self.encoder( x[:, :, 1 + 4 * (i - 1):1 + 4 * i, :, :], - feat_cache=self._enc_feat_map, - feat_idx=self._enc_conv_idx) + feat_cache=feat_map, + feat_idx=conv_idx) out = torch.cat([out, out_], 2) mu, log_var = self.conv1(out).chunk(2, dim=1) - self.clear_cache() return mu def decode(self, z): - self.clear_cache() + conv_idx = [0] + feat_map = [None] * count_conv3d(self.decoder) # z: [b,c,t,h,w] iter_ = z.shape[2] x = self.conv2(z) for i in range(iter_): - self._conv_idx = [0] + conv_idx = [0] if i == 0: out = self.decoder( x[:, :, i:i + 1, :, :], - feat_cache=self._feat_map, - feat_idx=self._conv_idx) + feat_cache=feat_map, + feat_idx=conv_idx) else: out_ = self.decoder( x[:, :, i:i + 1, :, :], - feat_cache=self._feat_map, - feat_idx=self._conv_idx) + feat_cache=feat_map, + feat_idx=conv_idx) out = torch.cat([out, out_], 2) - self.clear_cache() return out - - def clear_cache(self): - self._conv_num = count_conv3d(self.decoder) - self._conv_idx = [0] - self._feat_map = [None] * self._conv_num - #cache encode - self._enc_conv_num = count_conv3d(self.encoder) - self._enc_conv_idx = [0] - self._enc_feat_map = [None] * self._enc_conv_num From 0e9d1724be327c79ba86159d868f0b57adb8c384 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Wed, 1 Oct 2025 21:33:05 -0700 Subject: [PATCH 078/114] Add a .bat to the AMD portable to disable smart memory. (#10153) --- .ci/windows_amd_base_files/README_VERY_IMPORTANT.txt | 5 ++++- .../run_amd_gpu_disable_smart_memory.bat | 2 ++ 2 files changed, 6 insertions(+), 1 deletion(-) create mode 100755 .ci/windows_amd_base_files/run_amd_gpu_disable_smart_memory.bat diff --git a/.ci/windows_amd_base_files/README_VERY_IMPORTANT.txt b/.ci/windows_amd_base_files/README_VERY_IMPORTANT.txt index 570ac3398..96a500be2 100755 --- a/.ci/windows_amd_base_files/README_VERY_IMPORTANT.txt +++ b/.ci/windows_amd_base_files/README_VERY_IMPORTANT.txt @@ -3,10 +3,13 @@ https://www.amd.com/en/resources/support-articles/release-notes/RN-AMDGPU-WINDOW HOW TO RUN: -if you have a AMD gpu: +If you have a AMD gpu: run_amd_gpu.bat +If you have memory issues you can try disabling the smart memory management by running comfyui with: + +run_amd_gpu_disable_smart_memory.bat IF YOU GET A RED ERROR IN THE UI MAKE SURE YOU HAVE A MODEL/CHECKPOINT IN: ComfyUI\models\checkpoints diff --git a/.ci/windows_amd_base_files/run_amd_gpu_disable_smart_memory.bat b/.ci/windows_amd_base_files/run_amd_gpu_disable_smart_memory.bat new file mode 100755 index 000000000..cece0aeb2 --- /dev/null +++ b/.ci/windows_amd_base_files/run_amd_gpu_disable_smart_memory.bat @@ -0,0 +1,2 @@ +.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build --disable-smart-memory +pause From 8f4ee9984c0c3864290e4fea81cfea2ba281717d Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Thu, 2 Oct 2025 23:53:00 +0300 Subject: [PATCH 079/114] convert nodes_morphology.py to V3 schema (#10159) --- comfy_extras/nodes_morphology.py | 116 +++++++++++++++++++------------ 1 file changed, 70 insertions(+), 46 deletions(-) diff --git a/comfy_extras/nodes_morphology.py b/comfy_extras/nodes_morphology.py index 075b26c40..67377e1bc 100644 --- a/comfy_extras/nodes_morphology.py +++ b/comfy_extras/nodes_morphology.py @@ -1,24 +1,34 @@ import torch import comfy.model_management +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io from kornia.morphology import dilation, erosion, opening, closing, gradient, top_hat, bottom_hat import kornia.color -class Morphology: +class Morphology(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"image": ("IMAGE",), - "operation": (["erode", "dilate", "open", "close", "gradient", "bottom_hat", "top_hat"],), - "kernel_size": ("INT", {"default": 3, "min": 3, "max": 999, "step": 1}), - }} + def define_schema(cls): + return io.Schema( + node_id="Morphology", + display_name="ImageMorphology", + category="image/postprocessing", + inputs=[ + io.Image.Input("image"), + io.Combo.Input( + "operation", + options=["erode", "dilate", "open", "close", "gradient", "bottom_hat", "top_hat"], + ), + io.Int.Input("kernel_size", default=3, min=3, max=999, step=1), + ], + outputs=[ + io.Image.Output(), + ], + ) - RETURN_TYPES = ("IMAGE",) - FUNCTION = "process" - - CATEGORY = "image/postprocessing" - - def process(self, image, operation, kernel_size): + @classmethod + def execute(cls, image, operation, kernel_size) -> io.NodeOutput: device = comfy.model_management.get_torch_device() kernel = torch.ones(kernel_size, kernel_size, device=device) image_k = image.to(device).movedim(-1, 1) @@ -39,49 +49,63 @@ class Morphology: else: raise ValueError(f"Invalid operation {operation} for morphology. Must be one of 'erode', 'dilate', 'open', 'close', 'gradient', 'tophat', 'bottomhat'") img_out = output.to(comfy.model_management.intermediate_device()).movedim(1, -1) - return (img_out,) + return io.NodeOutput(img_out) -class ImageRGBToYUV: +class ImageRGBToYUV(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "image": ("IMAGE",), - }} + def define_schema(cls): + return io.Schema( + node_id="ImageRGBToYUV", + category="image/batch", + inputs=[ + io.Image.Input("image"), + ], + outputs=[ + io.Image.Output(display_name="Y"), + io.Image.Output(display_name="U"), + io.Image.Output(display_name="V"), + ], + ) - RETURN_TYPES = ("IMAGE", "IMAGE", "IMAGE") - RETURN_NAMES = ("Y", "U", "V") - FUNCTION = "execute" - - CATEGORY = "image/batch" - - def execute(self, image): + @classmethod + def execute(cls, image) -> io.NodeOutput: out = kornia.color.rgb_to_ycbcr(image.movedim(-1, 1)).movedim(1, -1) - return (out[..., 0:1].expand_as(image), out[..., 1:2].expand_as(image), out[..., 2:3].expand_as(image)) + return io.NodeOutput(out[..., 0:1].expand_as(image), out[..., 1:2].expand_as(image), out[..., 2:3].expand_as(image)) -class ImageYUVToRGB: +class ImageYUVToRGB(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"Y": ("IMAGE",), - "U": ("IMAGE",), - "V": ("IMAGE",), - }} + def define_schema(cls): + return io.Schema( + node_id="ImageYUVToRGB", + category="image/batch", + inputs=[ + io.Image.Input("Y"), + io.Image.Input("U"), + io.Image.Input("V"), + ], + outputs=[ + io.Image.Output(), + ], + ) - RETURN_TYPES = ("IMAGE",) - FUNCTION = "execute" - - CATEGORY = "image/batch" - - def execute(self, Y, U, V): + @classmethod + def execute(cls, Y, U, V) -> io.NodeOutput: image = torch.cat([torch.mean(Y, dim=-1, keepdim=True), torch.mean(U, dim=-1, keepdim=True), torch.mean(V, dim=-1, keepdim=True)], dim=-1) out = kornia.color.ycbcr_to_rgb(image.movedim(-1, 1)).movedim(1, -1) - return (out,) + return io.NodeOutput(out) -NODE_CLASS_MAPPINGS = { - "Morphology": Morphology, - "ImageRGBToYUV": ImageRGBToYUV, - "ImageYUVToRGB": ImageYUVToRGB, -} -NODE_DISPLAY_NAME_MAPPINGS = { - "Morphology": "ImageMorphology", -} +class MorphologyExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + Morphology, + ImageRGBToYUV, + ImageYUVToRGB, + ] + + +async def comfy_entrypoint() -> MorphologyExtension: + return MorphologyExtension() + From f6e3e9a456127a7e539929f42ea6cac838197879 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 00:50:31 +0300 Subject: [PATCH 080/114] fix(api-nodes): made logging path to be smaller (#10156) --- comfy_api_nodes/apis/client.py | 5 +- comfy_api_nodes/apis/request_logger.py | 72 ++++++++++++++++++++------ 2 files changed, 59 insertions(+), 18 deletions(-) diff --git a/comfy_api_nodes/apis/client.py b/comfy_api_nodes/apis/client.py index 0aed906fb..18a694675 100644 --- a/comfy_api_nodes/apis/client.py +++ b/comfy_api_nodes/apis/client.py @@ -95,6 +95,7 @@ import aiohttp import asyncio import logging import io +import os import socket from aiohttp.client_exceptions import ClientError, ClientResponseError from typing import Dict, Type, Optional, Any, TypeVar, Generic, Callable, Tuple @@ -499,7 +500,9 @@ class ApiClient: else: raise ValueError("File must be BytesIO or str path") - operation_id = f"upload_{upload_url.split('/')[-1]}_{uuid.uuid4().hex[:8]}" + parsed = urlparse(upload_url) + basename = os.path.basename(parsed.path) or parsed.netloc or "upload" + operation_id = f"upload_{basename}_{uuid.uuid4().hex[:8]}" request_logger.log_request_response( operation_id=operation_id, request_method="PUT", diff --git a/comfy_api_nodes/apis/request_logger.py b/comfy_api_nodes/apis/request_logger.py index 42901e141..2e0ca5380 100644 --- a/comfy_api_nodes/apis/request_logger.py +++ b/comfy_api_nodes/apis/request_logger.py @@ -4,16 +4,18 @@ import os import datetime import json import logging +import re +import hashlib +from typing import Any + import folder_paths # Get the logger instance logger = logging.getLogger(__name__) + def get_log_directory(): - """ - Ensures the API log directory exists within ComfyUI's temp directory - and returns its path. - """ + """Ensures the API log directory exists within ComfyUI's temp directory and returns its path.""" base_temp_dir = folder_paths.get_temp_directory() log_dir = os.path.join(base_temp_dir, "api_logs") try: @@ -24,42 +26,77 @@ def get_log_directory(): return base_temp_dir return log_dir -def _format_data_for_logging(data): + +def _sanitize_filename_component(name: str) -> str: + if not name: + return "log" + sanitized = re.sub(r"[^A-Za-z0-9._-]+", "_", name) # Replace disallowed characters with underscore + sanitized = sanitized.strip(" ._") # Windows: trailing dots or spaces are not allowed + if not sanitized: + sanitized = "log" + return sanitized + + +def _short_hash(*parts: str, length: int = 10) -> str: + return hashlib.sha1(("|".join(parts)).encode("utf-8")).hexdigest()[:length] + + +def _build_log_filepath(log_dir: str, operation_id: str, request_url: str) -> str: + """Build log filepath. We keep it well under common path length limits aiming for <= 240 characters total.""" + timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S_%f") + slug = _sanitize_filename_component(operation_id) # Best-effort human-readable slug from operation_id + h = _short_hash(operation_id or "", request_url or "") # Short hash ties log to the full operation and URL + + # Compute how much room we have for the slug given the directory length + # Keep total path length reasonably below ~260 on Windows. + max_total_path = 240 + prefix = f"{timestamp}_" + suffix = f"_{h}.log" + if not slug: + slug = "op" + max_filename_len = max(60, max_total_path - len(log_dir) - 1) + max_slug_len = max(8, max_filename_len - len(prefix) - len(suffix)) + if len(slug) > max_slug_len: + slug = slug[:max_slug_len].rstrip(" ._-") + return os.path.join(log_dir, f"{prefix}{slug}{suffix}") + + +def _format_data_for_logging(data: Any) -> str: """Helper to format data (dict, str, bytes) for logging.""" if isinstance(data, bytes): try: - return data.decode('utf-8') # Try to decode as text + return data.decode("utf-8") # Try to decode as text except UnicodeDecodeError: return f"[Binary data of length {len(data)} bytes]" elif isinstance(data, (dict, list)): try: return json.dumps(data, indent=2, ensure_ascii=False) except TypeError: - return str(data) # Fallback for non-serializable objects + return str(data) # Fallback for non-serializable objects return str(data) + def log_request_response( operation_id: str, request_method: str, request_url: str, request_headers: dict | None = None, request_params: dict | None = None, - request_data: any = None, + request_data: Any = None, response_status_code: int | None = None, response_headers: dict | None = None, - response_content: any = None, - error_message: str | None = None + response_content: Any = None, + error_message: str | None = None, ): """ Logs API request and response details to a file in the temp/api_logs directory. + Filenames are sanitized and length-limited for cross-platform safety. + If we still fail to write, we fall back to appending into api.log. """ log_dir = get_log_directory() - timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S_%f") - filename = f"{timestamp}_{operation_id.replace('/', '_').replace(':', '_')}.log" - filepath = os.path.join(log_dir, filename) - - log_content = [] + filepath = _build_log_filepath(log_dir, operation_id, request_url) + log_content: list[str] = [] log_content.append(f"Timestamp: {datetime.datetime.now().isoformat()}") log_content.append(f"Operation ID: {operation_id}") log_content.append("-" * 30 + " REQUEST " + "-" * 30) @@ -69,7 +106,7 @@ def log_request_response( log_content.append(f"Headers:\n{_format_data_for_logging(request_headers)}") if request_params: log_content.append(f"Params:\n{_format_data_for_logging(request_params)}") - if request_data: + if request_data is not None: log_content.append(f"Data/Body:\n{_format_data_for_logging(request_data)}") log_content.append("\n" + "-" * 30 + " RESPONSE " + "-" * 30) @@ -77,7 +114,7 @@ def log_request_response( log_content.append(f"Status Code: {response_status_code}") if response_headers: log_content.append(f"Headers:\n{_format_data_for_logging(response_headers)}") - if response_content: + if response_content is not None: log_content.append(f"Content:\n{_format_data_for_logging(response_content)}") if error_message: log_content.append(f"Error:\n{error_message}") @@ -89,6 +126,7 @@ def log_request_response( except Exception as e: logger.error(f"Error writing API log to {filepath}: {e}") + if __name__ == '__main__': # Example usage (for testing the logger directly) logger.setLevel(logging.DEBUG) From e9364ee279f65d0546fea1796c3cd2e0b7e1965f Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Thu, 2 Oct 2025 14:57:15 -0700 Subject: [PATCH 081/114] Turn on TORCH_ROCM_AOTRITON_ENABLE_EXPERIMENTAL by default. (#10168) --- main.py | 1 + 1 file changed, 1 insertion(+) diff --git a/main.py b/main.py index 70696fcc3..35857dba8 100644 --- a/main.py +++ b/main.py @@ -115,6 +115,7 @@ if os.name == "nt": os.environ['MIMALLOC_PURGE_DELAY'] = '0' if __name__ == "__main__": + os.environ['TORCH_ROCM_AOTRITON_ENABLE_EXPERIMENTAL'] = '1' if args.default_device is not None: default_dev = args.default_device devices = list(range(32)) From 1395bce9f707e52ec613eeaa87ea690518cfe0a8 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 01:20:29 +0300 Subject: [PATCH 082/114] update example_node to use V3 schema (#9723) --- custom_nodes/example_node.py.example | 161 +++++++++++---------------- 1 file changed, 68 insertions(+), 93 deletions(-) diff --git a/custom_nodes/example_node.py.example b/custom_nodes/example_node.py.example index 29ab2aa72..779c35787 100644 --- a/custom_nodes/example_node.py.example +++ b/custom_nodes/example_node.py.example @@ -1,96 +1,70 @@ -class Example: +from typing_extensions import override + +from comfy_api.latest import ComfyExtension, io + + +class Example(io.ComfyNode): """ - A example node + An example node Class methods ------------- - INPUT_TYPES (dict): - Tell the main program input parameters of nodes. - IS_CHANGED: + define_schema (io.Schema): + Tell the main program the metadata, input, output parameters of nodes. + fingerprint_inputs: optional method to control when the node is re executed. + check_lazy_status: + optional method to control list of input names that need to be evaluated. - Attributes - ---------- - RETURN_TYPES (`tuple`): - The type of each element in the output tuple. - RETURN_NAMES (`tuple`): - Optional: The name of each output in the output tuple. - FUNCTION (`str`): - The name of the entry-point method. For example, if `FUNCTION = "execute"` then it will run Example().execute() - OUTPUT_NODE ([`bool`]): - If this node is an output node that outputs a result/image from the graph. The SaveImage node is an example. - The backend iterates on these output nodes and tries to execute all their parents if their parent graph is properly connected. - Assumed to be False if not present. - CATEGORY (`str`): - The category the node should appear in the UI. - DEPRECATED (`bool`): - Indicates whether the node is deprecated. Deprecated nodes are hidden by default in the UI, but remain - functional in existing workflows that use them. - EXPERIMENTAL (`bool`): - Indicates whether the node is experimental. Experimental nodes are marked as such in the UI and may be subject to - significant changes or removal in future versions. Use with caution in production workflows. - execute(s) -> tuple || None: - The entry point method. The name of this method must be the same as the value of property `FUNCTION`. - For example, if `FUNCTION = "execute"` then this method's name must be `execute`, if `FUNCTION = "foo"` then it must be `foo`. """ - def __init__(self): - pass @classmethod - def INPUT_TYPES(s): + def define_schema(cls) -> io.Schema: """ - Return a dictionary which contains config for all input fields. - Some types (string): "MODEL", "VAE", "CLIP", "CONDITIONING", "LATENT", "IMAGE", "INT", "STRING", "FLOAT". - Input types "INT", "STRING" or "FLOAT" are special values for fields on the node. - The type can be a list for selection. - - Returns: `dict`: - - Key input_fields_group (`string`): Can be either required, hidden or optional. A node class must have property `required` - - Value input_fields (`dict`): Contains input fields config: - * Key field_name (`string`): Name of a entry-point method's argument - * Value field_config (`tuple`): - + First value is a string indicate the type of field or a list for selection. - + Second value is a config for type "INT", "STRING" or "FLOAT". + Return a schema which contains all information about the node. + Some types: "Model", "Vae", "Clip", "Conditioning", "Latent", "Image", "Int", "String", "Float", "Combo". + For outputs the "io.Model.Output" should be used, for inputs the "io.Model.Input" can be used. + The type can be a "Combo" - this will be a list for selection. """ - return { - "required": { - "image": ("IMAGE",), - "int_field": ("INT", { - "default": 0, - "min": 0, #Minimum value - "max": 4096, #Maximum value - "step": 64, #Slider's step - "display": "number", # Cosmetic only: display as "number" or "slider" - "lazy": True # Will only be evaluated if check_lazy_status requires it - }), - "float_field": ("FLOAT", { - "default": 1.0, - "min": 0.0, - "max": 10.0, - "step": 0.01, - "round": 0.001, #The value representing the precision to round to, will be set to the step value by default. Can be set to False to disable rounding. - "display": "number", - "lazy": True - }), - "print_to_screen": (["enable", "disable"],), - "string_field": ("STRING", { - "multiline": False, #True if you want the field to look like the one on the ClipTextEncode node - "default": "Hello World!", - "lazy": True - }), - }, - } + return io.Schema( + node_id="Example", + display_name="Example Node", + category="Example", + inputs=[ + io.Image.Input("image"), + io.Int.Input( + "int_field", + min=0, + max=4096, + step=64, # Slider's step + display_mode=io.NumberDisplay.number, # Cosmetic only: display as "number" or "slider" + lazy=True, # Will only be evaluated if check_lazy_status requires it + ), + io.Float.Input( + "float_field", + default=1.0, + min=0.0, + max=10.0, + step=0.01, + round=0.001, #The value representing the precision to round to, will be set to the step value by default. Can be set to False to disable rounding. + display_mode=io.NumberDisplay.number, + lazy=True, + ), + io.Combo.Input("print_to_screen", options=["enable", "disable"]), + io.String.Input( + "string_field", + multiline=False, # True if you want the field to look like the one on the ClipTextEncode node + default="Hello world!", + lazy=True, + ) + ], + outputs=[ + io.Image.Output(), + ], + ) - RETURN_TYPES = ("IMAGE",) - #RETURN_NAMES = ("image_output_name",) - - FUNCTION = "test" - - #OUTPUT_NODE = False - - CATEGORY = "Example" - - def check_lazy_status(self, image, string_field, int_field, float_field, print_to_screen): + @classmethod + def check_lazy_status(cls, image, string_field, int_field, float_field, print_to_screen): """ Return a list of input names that need to be evaluated. @@ -107,7 +81,8 @@ class Example: else: return [] - def test(self, image, string_field, int_field, float_field, print_to_screen): + @classmethod + def execute(cls, image, string_field, int_field, float_field, print_to_screen) -> io.NodeOutput: if print_to_screen == "enable": print(f"""Your input contains: string_field aka input text: {string_field} @@ -116,7 +91,7 @@ class Example: """) #do some processing on the image, in this example I just invert it image = 1.0 - image - return (image,) + return io.NodeOutput(image) """ The node will always be re executed if any of the inputs change but @@ -127,7 +102,7 @@ class Example: changes between executions the LoadImage node is executed again. """ #@classmethod - #def IS_CHANGED(s, image, string_field, int_field, float_field, print_to_screen): + #def fingerprint_inputs(s, image, string_field, int_field, float_field, print_to_screen): # return "" # Set the web directory, any .js file in that directory will be loaded by the frontend as a frontend extension @@ -143,13 +118,13 @@ async def get_hello(request): return web.json_response("hello") -# A dictionary that contains all nodes you want to export with their names -# NOTE: names should be globally unique -NODE_CLASS_MAPPINGS = { - "Example": Example -} +class ExampleExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + Example, + ] -# A dictionary that contains the friendly/humanly readable titles for the nodes -NODE_DISPLAY_NAME_MAPPINGS = { - "Example": "Example Node" -} + +async def comfy_entrypoint() -> ExampleExtension: # ComfyUI calls this to load your extension and its nodes. + return ExampleExtension() From 4ffea0e864275301329ddb5ecc3fbc7211d7a802 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 02:14:28 +0300 Subject: [PATCH 083/114] feat(linter, api-nodes): add pylint for comfy_api_nodes folder (#10157) --- .github/workflows/ruff.yml | 25 ++++++++++++++ comfy_api_nodes/apis/__init__.py | 1 + comfy_api_nodes/apis/client.py | 2 +- comfy_api_nodes/apis/rodin_api.py | 4 --- pyproject.toml | 54 +++++++++++++++++++++++++++++++ 5 files changed, 81 insertions(+), 5 deletions(-) diff --git a/.github/workflows/ruff.yml b/.github/workflows/ruff.yml index 4c1a02594..b24d86a6b 100644 --- a/.github/workflows/ruff.yml +++ b/.github/workflows/ruff.yml @@ -21,3 +21,28 @@ jobs: - name: Run Ruff run: ruff check . + + pylint: + name: Run Pylint + runs-on: ubuntu-latest + + steps: + - name: Checkout repository + uses: actions/checkout@v4 + + - name: Set up Python + uses: actions/setup-python@v4 + with: + python-version: '3.12' + + - name: Install requirements + run: | + python -m pip install --upgrade pip + pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu + pip install -r requirements.txt + + - name: Install Pylint + run: pip install pylint + + - name: Run Pylint + run: pylint comfy_api_nodes diff --git a/comfy_api_nodes/apis/__init__.py b/comfy_api_nodes/apis/__init__.py index 78a23db30..98f9e540d 100644 --- a/comfy_api_nodes/apis/__init__.py +++ b/comfy_api_nodes/apis/__init__.py @@ -2,6 +2,7 @@ # filename: filtered-openapi.yaml # timestamp: 2025-07-30T08:54:00+00:00 +# pylint: disable from __future__ import annotations from datetime import date, datetime diff --git a/comfy_api_nodes/apis/client.py b/comfy_api_nodes/apis/client.py index 18a694675..79de3c262 100644 --- a/comfy_api_nodes/apis/client.py +++ b/comfy_api_nodes/apis/client.py @@ -535,7 +535,7 @@ class ApiClient: request_method="PUT", request_url=upload_url, response_status_code=e.status if hasattr(e, "status") else None, - response_headers=dict(e.headers) if getattr(e, "headers") else None, + response_headers=dict(e.headers) if hasattr(e, "headers") else None, response_content=None, error_message=f"{type(e).__name__}: {str(e)}", ) diff --git a/comfy_api_nodes/apis/rodin_api.py b/comfy_api_nodes/apis/rodin_api.py index 02cf42c29..fc26a6e73 100644 --- a/comfy_api_nodes/apis/rodin_api.py +++ b/comfy_api_nodes/apis/rodin_api.py @@ -52,7 +52,3 @@ class RodinResourceItem(BaseModel): class Rodin3DDownloadResponse(BaseModel): list: List[RodinResourceItem] = Field(..., description="Source List") - - - - diff --git a/pyproject.toml b/pyproject.toml index d0a76c6d0..598af4157 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -22,3 +22,57 @@ lint.select = [ "F", ] exclude = ["*.ipynb", "**/generated/*.pyi"] + +[tool.pylint] +master.py-version = "3.9" +master.extension-pkg-allow-list = [ + "pydantic", +] +reports.output-format = "colorized" +similarities.ignore-imports = "yes" +messages_control.disable = [ + "missing-module-docstring", + "missing-class-docstring", + "missing-function-docstring", + "line-too-long", + "too-few-public-methods", + "too-many-public-methods", + "too-many-instance-attributes", + "too-many-positional-arguments", + "broad-exception-raised", + "too-many-lines", + "invalid-name", + "unused-argument", + "broad-exception-caught", + "consider-using-with", + "fixme", + "too-many-statements", + "too-many-branches", + "too-many-locals", + "too-many-arguments", + "duplicate-code", + "abstract-method", + "superfluous-parens", + "arguments-differ", + "redefined-builtin", + "unnecessary-lambda", + "dangerous-default-value", + # next warnings should be fixed in future + "bad-classmethod-argument", # Class method should have 'cls' as first argument + "wrong-import-order", # Standard imports should be placed before third party imports + "logging-fstring-interpolation", # Use lazy % formatting in logging functions + "ungrouped-imports", + "unnecessary-pass", + "unidiomatic-typecheck", + "unnecessary-lambda-assignment", + "bad-indentation", + "no-else-return", + "no-else-raise", + "invalid-overridden-method", + "unused-variable", + "pointless-string-statement", + "inconsistent-return-statements", + "import-outside-toplevel", + "reimported", + "redefined-outer-name", +] From ed3ca78e080d697b6cf29497c07e14ee9c27a3ac Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 21:26:34 +0300 Subject: [PATCH 084/114] feat(api-nodes): add kling-2-5-turbo to txt2video and img2video nodes (#10155) --- comfy_api_nodes/apis/__init__.py | 2 ++ comfy_api_nodes/nodes_kling.py | 2 ++ 2 files changed, 4 insertions(+) diff --git a/comfy_api_nodes/apis/__init__.py b/comfy_api_nodes/apis/__init__.py index 98f9e540d..ee2aa1ce6 100644 --- a/comfy_api_nodes/apis/__init__.py +++ b/comfy_api_nodes/apis/__init__.py @@ -1321,6 +1321,7 @@ class KlingTextToVideoModelName(str, Enum): kling_v1 = 'kling-v1' kling_v1_6 = 'kling-v1-6' kling_v2_1_master = 'kling-v2-1-master' + kling_v2_5_turbo = 'kling-v2-5-turbo' class KlingVideoGenAspectRatio(str, Enum): @@ -1355,6 +1356,7 @@ class KlingVideoGenModelName(str, Enum): kling_v2_master = 'kling-v2-master' kling_v2_1 = 'kling-v2-1' kling_v2_1_master = 'kling-v2-1-master' + kling_v2_5_turbo = 'kling-v2-5-turbo' class KlingVideoResult(BaseModel): diff --git a/comfy_api_nodes/nodes_kling.py b/comfy_api_nodes/nodes_kling.py index 5f55b2cc9..d8646f106 100644 --- a/comfy_api_nodes/nodes_kling.py +++ b/comfy_api_nodes/nodes_kling.py @@ -423,6 +423,8 @@ class KlingTextToVideoNode(KlingNodeBase): "standard mode / 10s duration / kling-v2-master": ("std", "10", "kling-v2-master"), "pro mode / 5s duration / kling-v2-1-master": ("pro", "5", "kling-v2-1-master"), "pro mode / 10s duration / kling-v2-1-master": ("pro", "10", "kling-v2-1-master"), + "pro mode / 5s duration / kling-v2-5-turbo": ("pro", "5", "kling-v2-5-turbo"), + "pro mode / 10s duration / kling-v2-5-turbo": ("pro", "10", "kling-v2-5-turbo"), } @classmethod From 8a293372ecdea0ff8647921eaf3bb10c3d992abf Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 21:40:27 +0300 Subject: [PATCH 085/114] fix(api-nodes): reimport of base64 in Gemini node (#10181) --- comfy_api_nodes/nodes_gemini.py | 1 - pyproject.toml | 1 - 2 files changed, 2 deletions(-) diff --git a/comfy_api_nodes/nodes_gemini.py b/comfy_api_nodes/nodes_gemini.py index baa379b75..151cb4044 100644 --- a/comfy_api_nodes/nodes_gemini.py +++ b/comfy_api_nodes/nodes_gemini.py @@ -490,7 +490,6 @@ class GeminiInputFiles(ComfyNodeABC): # Use base64 string directly, not the data URI with open(file_path, "rb") as f: file_content = f.read() - import base64 base64_str = base64.b64encode(file_content).decode("utf-8") return GeminiPart( diff --git a/pyproject.toml b/pyproject.toml index 598af4157..7952f7f37 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -73,6 +73,5 @@ messages_control.disable = [ "pointless-string-statement", "inconsistent-return-statements", "import-outside-toplevel", - "reimported", "redefined-outer-name", ] From c2c5a7d5f80579bb44c11de0ce6eff94d1c111b9 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 21:41:06 +0300 Subject: [PATCH 086/114] fix(api-nodes): bad indentation in Recraft API node function (#10175) --- comfy_api_nodes/nodes_recraft.py | 78 ++++++++++++++++---------------- pyproject.toml | 1 - 2 files changed, 39 insertions(+), 40 deletions(-) diff --git a/comfy_api_nodes/nodes_recraft.py b/comfy_api_nodes/nodes_recraft.py index c8516b368..a006104b7 100644 --- a/comfy_api_nodes/nodes_recraft.py +++ b/comfy_api_nodes/nodes_recraft.py @@ -38,48 +38,48 @@ from PIL import UnidentifiedImageError async def handle_recraft_file_request( - image: torch.Tensor, - path: str, - mask: torch.Tensor=None, - total_pixels=4096*4096, - timeout=1024, - request=None, - auth_kwargs: dict[str,str] = None, - ) -> list[BytesIO]: - """ - Handle sending common Recraft file-only request to get back file bytes. - """ - if request is None: - request = EmptyRequest() + image: torch.Tensor, + path: str, + mask: torch.Tensor=None, + total_pixels=4096*4096, + timeout=1024, + request=None, + auth_kwargs: dict[str,str] = None, +) -> list[BytesIO]: + """ + Handle sending common Recraft file-only request to get back file bytes. + """ + if request is None: + request = EmptyRequest() - files = { - 'image': tensor_to_bytesio(image, total_pixels=total_pixels).read() - } - if mask is not None: - files['mask'] = tensor_to_bytesio(mask, total_pixels=total_pixels).read() + files = { + 'image': tensor_to_bytesio(image, total_pixels=total_pixels).read() + } + if mask is not None: + files['mask'] = tensor_to_bytesio(mask, total_pixels=total_pixels).read() - operation = SynchronousOperation( - endpoint=ApiEndpoint( - path=path, - method=HttpMethod.POST, - request_model=type(request), - response_model=RecraftImageGenerationResponse, - ), - request=request, - files=files, - content_type="multipart/form-data", - auth_kwargs=auth_kwargs, - multipart_parser=recraft_multipart_parser, - ) - response: RecraftImageGenerationResponse = await operation.execute() - all_bytesio = [] - if response.image is not None: - all_bytesio.append(await download_url_to_bytesio(response.image.url, timeout=timeout)) - else: - for data in response.data: - all_bytesio.append(await download_url_to_bytesio(data.url, timeout=timeout)) + operation = SynchronousOperation( + endpoint=ApiEndpoint( + path=path, + method=HttpMethod.POST, + request_model=type(request), + response_model=RecraftImageGenerationResponse, + ), + request=request, + files=files, + content_type="multipart/form-data", + auth_kwargs=auth_kwargs, + multipart_parser=recraft_multipart_parser, + ) + response: RecraftImageGenerationResponse = await operation.execute() + all_bytesio = [] + if response.image is not None: + all_bytesio.append(await download_url_to_bytesio(response.image.url, timeout=timeout)) + else: + for data in response.data: + all_bytesio.append(await download_url_to_bytesio(data.url, timeout=timeout)) - return all_bytesio + return all_bytesio def recraft_multipart_parser(data, parent_key=None, formatter: callable=None, converted_to_check: list[list]=None, is_list=False) -> dict: diff --git a/pyproject.toml b/pyproject.toml index 7952f7f37..240919a43 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -65,7 +65,6 @@ messages_control.disable = [ "unnecessary-pass", "unidiomatic-typecheck", "unnecessary-lambda-assignment", - "bad-indentation", "no-else-return", "no-else-raise", "invalid-overridden-method", From 3e68bc342cd60b909b4117c1b68a3afc62ef875c Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 21:43:54 +0300 Subject: [PATCH 087/114] convert nodes_torch_compile.py to V3 schema (#10173) --- comfy_extras/nodes_torch_compile.py | 46 +++++++++++++++++++---------- 1 file changed, 31 insertions(+), 15 deletions(-) diff --git a/comfy_extras/nodes_torch_compile.py b/comfy_extras/nodes_torch_compile.py index 605536678..adbeece2f 100644 --- a/comfy_extras/nodes_torch_compile.py +++ b/comfy_extras/nodes_torch_compile.py @@ -1,23 +1,39 @@ +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io from comfy_api.torch_helpers import set_torch_compile_wrapper -class TorchCompileModel: +class TorchCompileModel(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "model": ("MODEL",), - "backend": (["inductor", "cudagraphs"],), - }} - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" + def define_schema(cls) -> io.Schema: + return io.Schema( + node_id="TorchCompileModel", + category="_for_testing", + inputs=[ + io.Model.Input("model"), + io.Combo.Input( + "backend", + options=["inductor", "cudagraphs"], + ), + ], + outputs=[io.Model.Output()], + is_experimental=True, + ) - CATEGORY = "_for_testing" - EXPERIMENTAL = True - - def patch(self, model, backend): + @classmethod + def execute(cls, model, backend) -> io.NodeOutput: m = model.clone() set_torch_compile_wrapper(model=m, backend=backend) - return (m, ) + return io.NodeOutput(m) -NODE_CLASS_MAPPINGS = { - "TorchCompileModel": TorchCompileModel, -} + +class TorchCompileExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + TorchCompileModel, + ] + + +async def comfy_entrypoint() -> TorchCompileExtension: + return TorchCompileExtension() From d7aa414141f02a456801704a3da323fa2ed8f5cc Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 21:45:02 +0300 Subject: [PATCH 088/114] convert nodes_eps.py to V3 schema (#10172) --- comfy_extras/nodes_eps.py | 62 ++++++++++++++++++++++++--------------- 1 file changed, 38 insertions(+), 24 deletions(-) diff --git a/comfy_extras/nodes_eps.py b/comfy_extras/nodes_eps.py index c8818f096..7852d85e5 100644 --- a/comfy_extras/nodes_eps.py +++ b/comfy_extras/nodes_eps.py @@ -1,4 +1,9 @@ -class EpsilonScaling: +from typing_extensions import override + +from comfy_api.latest import ComfyExtension, io + + +class EpsilonScaling(io.ComfyNode): """ Implements the Epsilon Scaling method from 'Elucidating the Exposure Bias in Diffusion Models' (https://arxiv.org/abs/2308.15321v6). @@ -8,26 +13,28 @@ class EpsilonScaling: recommended by the paper for its practicality and effectiveness. """ @classmethod - def INPUT_TYPES(s): - return { - "required": { - "model": ("MODEL",), - "scaling_factor": ("FLOAT", { - "default": 1.005, - "min": 0.5, - "max": 1.5, - "step": 0.001, - "display": "number" - }), - } - } + def define_schema(cls): + return io.Schema( + node_id="Epsilon Scaling", + category="model_patches/unet", + inputs=[ + io.Model.Input("model"), + io.Float.Input( + "scaling_factor", + default=1.005, + min=0.5, + max=1.5, + step=0.001, + display_mode=io.NumberDisplay.number, + ), + ], + outputs=[ + io.Model.Output(), + ], + ) - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" - - CATEGORY = "model_patches/unet" - - def patch(self, model, scaling_factor): + @classmethod + def execute(cls, model, scaling_factor) -> io.NodeOutput: # Prevent division by zero, though the UI's min value should prevent this. if scaling_factor == 0: scaling_factor = 1e-9 @@ -53,8 +60,15 @@ class EpsilonScaling: model_clone.set_model_sampler_post_cfg_function(epsilon_scaling_function) - return (model_clone,) + return io.NodeOutput(model_clone) -NODE_CLASS_MAPPINGS = { - "Epsilon Scaling": EpsilonScaling -} + +class EpsilonScalingExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + EpsilonScaling, + ] + +async def comfy_entrypoint() -> EpsilonScalingExtension: + return EpsilonScalingExtension() From 8c26d7bbe6663f589f0a9562921aafb3c48955c6 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 21:48:21 +0300 Subject: [PATCH 089/114] convert nodes_pixverse.py to V3 schema (#10177) --- comfy_api_nodes/nodes_pixverse.py | 471 +++++++++++++++--------------- 1 file changed, 238 insertions(+), 233 deletions(-) diff --git a/comfy_api_nodes/nodes_pixverse.py b/comfy_api_nodes/nodes_pixverse.py index 7c5a52feb..eb98e9653 100644 --- a/comfy_api_nodes/nodes_pixverse.py +++ b/comfy_api_nodes/nodes_pixverse.py @@ -1,5 +1,7 @@ from inspect import cleandoc from typing import Optional +from typing_extensions import override +from io import BytesIO from comfy_api_nodes.apis.pixverse_api import ( PixverseTextVideoRequest, PixverseImageVideoRequest, @@ -26,12 +28,11 @@ from comfy_api_nodes.apinode_utils import ( tensor_to_bytesio, validate_string, ) -from comfy.comfy_types.node_typing import IO, ComfyNodeABC from comfy_api.input_impl import VideoFromFile +from comfy_api.latest import ComfyExtension, io as comfy_io import torch import aiohttp -from io import BytesIO AVERAGE_DURATION_T2V = 32 @@ -72,100 +73,101 @@ async def upload_image_to_pixverse(image: torch.Tensor, auth_kwargs=None): return response_upload.Resp.img_id -class PixverseTemplateNode: +class PixverseTemplateNode(comfy_io.ComfyNode): """ Select template for PixVerse Video generation. """ - RETURN_TYPES = (PixverseIO.TEMPLATE,) - RETURN_NAMES = ("pixverse_template",) - FUNCTION = "create_template" - CATEGORY = "api node/video/PixVerse" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="PixverseTemplateNode", + display_name="PixVerse Template", + category="api node/video/PixVerse", + inputs=[ + comfy_io.Combo.Input("template", options=[list(pixverse_templates.keys())]), + ], + outputs=[comfy_io.Custom(PixverseIO.TEMPLATE).Output(display_name="pixverse_template")], + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "template": (list(pixverse_templates.keys()),), - } - } - - def create_template(self, template: str): + def execute(cls, template: str) -> comfy_io.NodeOutput: template_id = pixverse_templates.get(template, None) if template_id is None: raise Exception(f"Template '{template}' is not recognized.") # just return the integer - return (template_id,) + return comfy_io.NodeOutput(template_id) -class PixverseTextToVideoNode(ComfyNodeABC): +class PixverseTextToVideoNode(comfy_io.ComfyNode): """ Generates videos based on prompt and output_size. """ - RETURN_TYPES = (IO.VIDEO,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/video/PixVerse" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="PixverseTextToVideoNode", + display_name="PixVerse Text to Video", + category="api node/video/PixVerse", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the video generation", + ), + comfy_io.Combo.Input( + "aspect_ratio", + options=[ratio.value for ratio in PixverseAspectRatio], + ), + comfy_io.Combo.Input( + "quality", + options=[resolution.value for resolution in PixverseQuality], + default=PixverseQuality.res_540p, + ), + comfy_io.Combo.Input( + "duration_seconds", + options=[dur.value for dur in PixverseDuration], + ), + comfy_io.Combo.Input( + "motion_mode", + options=[mode.value for mode in PixverseMotionMode], + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=2147483647, + control_after_generate=True, + tooltip="Seed for video generation.", + ), + comfy_io.String.Input( + "negative_prompt", + default="", + force_input=True, + tooltip="An optional text description of undesired elements on an image.", + optional=True, + ), + comfy_io.Custom(PixverseIO.TEMPLATE).Input( + "pixverse_template", + tooltip="An optional template to influence style of generation, created by the PixVerse Template node.", + optional=True, + ), + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the video generation", - }, - ), - "aspect_ratio": ([ratio.value for ratio in PixverseAspectRatio],), - "quality": ( - [resolution.value for resolution in PixverseQuality], - { - "default": PixverseQuality.res_540p, - }, - ), - "duration_seconds": ([dur.value for dur in PixverseDuration],), - "motion_mode": ([mode.value for mode in PixverseMotionMode],), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 2147483647, - "control_after_generate": True, - "tooltip": "Seed for video generation.", - }, - ), - }, - "optional": { - "negative_prompt": ( - IO.STRING, - { - "default": "", - "forceInput": True, - "tooltip": "An optional text description of undesired elements on an image.", - }, - ), - "pixverse_template": ( - PixverseIO.TEMPLATE, - { - "tooltip": "An optional template to influence style of generation, created by the PixVerse Template node." - }, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - async def api_call( - self, + async def execute( + cls, prompt: str, aspect_ratio: str, quality: str, @@ -174,9 +176,7 @@ class PixverseTextToVideoNode(ComfyNodeABC): seed, negative_prompt: str = None, pixverse_template: int = None, - unique_id: Optional[str] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: validate_string(prompt, strip_whitespace=False) # 1080p is limited to 5 seconds duration # only normal motion_mode supported for 1080p or for non-5 second duration @@ -186,6 +186,10 @@ class PixverseTextToVideoNode(ComfyNodeABC): elif duration_seconds != PixverseDuration.dur_5: motion_mode = PixverseMotionMode.normal + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } operation = SynchronousOperation( endpoint=ApiEndpoint( path="/proxy/pixverse/video/text/generate", @@ -203,7 +207,7 @@ class PixverseTextToVideoNode(ComfyNodeABC): template_id=pixverse_template, seed=seed, ), - auth_kwargs=kwargs, + auth_kwargs=auth, ) response_api = await operation.execute() @@ -224,8 +228,8 @@ class PixverseTextToVideoNode(ComfyNodeABC): PixverseStatus.deleted, ], status_extractor=lambda x: x.Resp.status, - auth_kwargs=kwargs, - node_id=unique_id, + auth_kwargs=auth, + node_id=cls.hidden.unique_id, result_url_extractor=get_video_url_from_response, estimated_duration=AVERAGE_DURATION_T2V, ) @@ -233,77 +237,75 @@ class PixverseTextToVideoNode(ComfyNodeABC): async with aiohttp.ClientSession() as session: async with session.get(response_poll.Resp.url) as vid_response: - return (VideoFromFile(BytesIO(await vid_response.content.read())),) + return comfy_io.NodeOutput(VideoFromFile(BytesIO(await vid_response.content.read()))) -class PixverseImageToVideoNode(ComfyNodeABC): +class PixverseImageToVideoNode(comfy_io.ComfyNode): """ Generates videos based on prompt and output_size. """ - RETURN_TYPES = (IO.VIDEO,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/video/PixVerse" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="PixverseImageToVideoNode", + display_name="PixVerse Image to Video", + category="api node/video/PixVerse", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("image"), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the video generation", + ), + comfy_io.Combo.Input( + "quality", + options=[resolution.value for resolution in PixverseQuality], + default=PixverseQuality.res_540p, + ), + comfy_io.Combo.Input( + "duration_seconds", + options=[dur.value for dur in PixverseDuration], + ), + comfy_io.Combo.Input( + "motion_mode", + options=[mode.value for mode in PixverseMotionMode], + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=2147483647, + control_after_generate=True, + tooltip="Seed for video generation.", + ), + comfy_io.String.Input( + "negative_prompt", + default="", + force_input=True, + tooltip="An optional text description of undesired elements on an image.", + optional=True, + ), + comfy_io.Custom(PixverseIO.TEMPLATE).Input( + "pixverse_template", + tooltip="An optional template to influence style of generation, created by the PixVerse Template node.", + optional=True, + ), + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": (IO.IMAGE,), - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the video generation", - }, - ), - "quality": ( - [resolution.value for resolution in PixverseQuality], - { - "default": PixverseQuality.res_540p, - }, - ), - "duration_seconds": ([dur.value for dur in PixverseDuration],), - "motion_mode": ([mode.value for mode in PixverseMotionMode],), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 2147483647, - "control_after_generate": True, - "tooltip": "Seed for video generation.", - }, - ), - }, - "optional": { - "negative_prompt": ( - IO.STRING, - { - "default": "", - "forceInput": True, - "tooltip": "An optional text description of undesired elements on an image.", - }, - ), - "pixverse_template": ( - PixverseIO.TEMPLATE, - { - "tooltip": "An optional template to influence style of generation, created by the PixVerse Template node." - }, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - async def api_call( - self, + async def execute( + cls, image: torch.Tensor, prompt: str, quality: str, @@ -312,11 +314,13 @@ class PixverseImageToVideoNode(ComfyNodeABC): seed, negative_prompt: str = None, pixverse_template: int = None, - unique_id: Optional[str] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: validate_string(prompt, strip_whitespace=False) - img_id = await upload_image_to_pixverse(image, auth_kwargs=kwargs) + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } + img_id = await upload_image_to_pixverse(image, auth_kwargs=auth) # 1080p is limited to 5 seconds duration # only normal motion_mode supported for 1080p or for non-5 second duration @@ -343,7 +347,7 @@ class PixverseImageToVideoNode(ComfyNodeABC): template_id=pixverse_template, seed=seed, ), - auth_kwargs=kwargs, + auth_kwargs=auth, ) response_api = await operation.execute() @@ -364,8 +368,8 @@ class PixverseImageToVideoNode(ComfyNodeABC): PixverseStatus.deleted, ], status_extractor=lambda x: x.Resp.status, - auth_kwargs=kwargs, - node_id=unique_id, + auth_kwargs=auth, + node_id=cls.hidden.unique_id, result_url_extractor=get_video_url_from_response, estimated_duration=AVERAGE_DURATION_I2V, ) @@ -373,72 +377,71 @@ class PixverseImageToVideoNode(ComfyNodeABC): async with aiohttp.ClientSession() as session: async with session.get(response_poll.Resp.url) as vid_response: - return (VideoFromFile(BytesIO(await vid_response.content.read())),) + return comfy_io.NodeOutput(VideoFromFile(BytesIO(await vid_response.content.read()))) -class PixverseTransitionVideoNode(ComfyNodeABC): +class PixverseTransitionVideoNode(comfy_io.ComfyNode): """ Generates videos based on prompt and output_size. """ - RETURN_TYPES = (IO.VIDEO,) - DESCRIPTION = cleandoc(__doc__ or "") # Handle potential None value - FUNCTION = "api_call" - API_NODE = True - CATEGORY = "api node/video/PixVerse" + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="PixverseTransitionVideoNode", + display_name="PixVerse Transition Video", + category="api node/video/PixVerse", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("first_frame"), + comfy_io.Image.Input("last_frame"), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Prompt for the video generation", + ), + comfy_io.Combo.Input( + "quality", + options=[resolution.value for resolution in PixverseQuality], + default=PixverseQuality.res_540p, + ), + comfy_io.Combo.Input( + "duration_seconds", + options=[dur.value for dur in PixverseDuration], + ), + comfy_io.Combo.Input( + "motion_mode", + options=[mode.value for mode in PixverseMotionMode], + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=2147483647, + control_after_generate=True, + tooltip="Seed for video generation.", + ), + comfy_io.String.Input( + "negative_prompt", + default="", + force_input=True, + tooltip="An optional text description of undesired elements on an image.", + optional=True, + ), + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - return { - "required": { - "first_frame": (IO.IMAGE,), - "last_frame": (IO.IMAGE,), - "prompt": ( - IO.STRING, - { - "multiline": True, - "default": "", - "tooltip": "Prompt for the video generation", - }, - ), - "quality": ( - [resolution.value for resolution in PixverseQuality], - { - "default": PixverseQuality.res_540p, - }, - ), - "duration_seconds": ([dur.value for dur in PixverseDuration],), - "motion_mode": ([mode.value for mode in PixverseMotionMode],), - "seed": ( - IO.INT, - { - "default": 0, - "min": 0, - "max": 2147483647, - "control_after_generate": True, - "tooltip": "Seed for video generation.", - }, - ), - }, - "optional": { - "negative_prompt": ( - IO.STRING, - { - "default": "", - "forceInput": True, - "tooltip": "An optional text description of undesired elements on an image.", - }, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - async def api_call( - self, + async def execute( + cls, first_frame: torch.Tensor, last_frame: torch.Tensor, prompt: str, @@ -447,12 +450,14 @@ class PixverseTransitionVideoNode(ComfyNodeABC): motion_mode: str, seed, negative_prompt: str = None, - unique_id: Optional[str] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: validate_string(prompt, strip_whitespace=False) - first_frame_id = await upload_image_to_pixverse(first_frame, auth_kwargs=kwargs) - last_frame_id = await upload_image_to_pixverse(last_frame, auth_kwargs=kwargs) + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } + first_frame_id = await upload_image_to_pixverse(first_frame, auth_kwargs=auth) + last_frame_id = await upload_image_to_pixverse(last_frame, auth_kwargs=auth) # 1080p is limited to 5 seconds duration # only normal motion_mode supported for 1080p or for non-5 second duration @@ -479,7 +484,7 @@ class PixverseTransitionVideoNode(ComfyNodeABC): negative_prompt=negative_prompt if negative_prompt else None, seed=seed, ), - auth_kwargs=kwargs, + auth_kwargs=auth, ) response_api = await operation.execute() @@ -500,8 +505,8 @@ class PixverseTransitionVideoNode(ComfyNodeABC): PixverseStatus.deleted, ], status_extractor=lambda x: x.Resp.status, - auth_kwargs=kwargs, - node_id=unique_id, + auth_kwargs=auth, + node_id=cls.hidden.unique_id, result_url_extractor=get_video_url_from_response, estimated_duration=AVERAGE_DURATION_T2V, ) @@ -509,19 +514,19 @@ class PixverseTransitionVideoNode(ComfyNodeABC): async with aiohttp.ClientSession() as session: async with session.get(response_poll.Resp.url) as vid_response: - return (VideoFromFile(BytesIO(await vid_response.content.read())),) + return comfy_io.NodeOutput(VideoFromFile(BytesIO(await vid_response.content.read()))) -NODE_CLASS_MAPPINGS = { - "PixverseTextToVideoNode": PixverseTextToVideoNode, - "PixverseImageToVideoNode": PixverseImageToVideoNode, - "PixverseTransitionVideoNode": PixverseTransitionVideoNode, - "PixverseTemplateNode": PixverseTemplateNode, -} +class PixVerseExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: + return [ + PixverseTextToVideoNode, + PixverseImageToVideoNode, + PixverseTransitionVideoNode, + PixverseTemplateNode, + ] -NODE_DISPLAY_NAME_MAPPINGS = { - "PixverseTextToVideoNode": "PixVerse Text to Video", - "PixverseImageToVideoNode": "PixVerse Image to Video", - "PixverseTransitionVideoNode": "PixVerse Transition Video", - "PixverseTemplateNode": "PixVerse Template", -} + +async def comfy_entrypoint() -> PixVerseExtension: + return PixVerseExtension() From 5c8e986e273d8af8b976fddbaed726e8278cf1fe Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 21:50:38 +0300 Subject: [PATCH 090/114] convert nodes_tomesd.py to V3 schema (#10180) --- comfy_extras/nodes_tomesd.py | 50 +++++++++++++++++++++++------------- 1 file changed, 32 insertions(+), 18 deletions(-) diff --git a/comfy_extras/nodes_tomesd.py b/comfy_extras/nodes_tomesd.py index 9f77c06fc..87bf29b8f 100644 --- a/comfy_extras/nodes_tomesd.py +++ b/comfy_extras/nodes_tomesd.py @@ -1,7 +1,9 @@ #Taken from: https://github.com/dbolya/tomesd import torch -from typing import Tuple, Callable +from typing import Tuple, Callable, Optional +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io import math def do_nothing(x: torch.Tensor, mode:str=None): @@ -144,33 +146,45 @@ def get_functions(x, ratio, original_shape): -class TomePatchModel: +class TomePatchModel(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "model": ("MODEL",), - "ratio": ("FLOAT", {"default": 0.3, "min": 0.0, "max": 1.0, "step": 0.01}), - }} - RETURN_TYPES = ("MODEL",) - FUNCTION = "patch" + def define_schema(cls): + return io.Schema( + node_id="TomePatchModel", + category="model_patches/unet", + inputs=[ + io.Model.Input("model"), + io.Float.Input("ratio", default=0.3, min=0.0, max=1.0, step=0.01), + ], + outputs=[io.Model.Output()], + ) - CATEGORY = "model_patches/unet" - - def patch(self, model, ratio): - self.u = None + @classmethod + def execute(cls, model, ratio) -> io.NodeOutput: + u: Optional[Callable] = None def tomesd_m(q, k, v, extra_options): + nonlocal u #NOTE: In the reference code get_functions takes x (input of the transformer block) as the argument instead of q #however from my basic testing it seems that using q instead gives better results - m, self.u = get_functions(q, ratio, extra_options["original_shape"]) + m, u = get_functions(q, ratio, extra_options["original_shape"]) return m(q), k, v def tomesd_u(n, extra_options): - return self.u(n) + nonlocal u + return u(n) m = model.clone() m.set_model_attn1_patch(tomesd_m) m.set_model_attn1_output_patch(tomesd_u) - return (m, ) + return io.NodeOutput(m) -NODE_CLASS_MAPPINGS = { - "TomePatchModel": TomePatchModel, -} +class TomePatchModelExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + TomePatchModel, + ] + + +async def comfy_entrypoint() -> TomePatchModelExtension: + return TomePatchModelExtension() From 4614ee09ca1aaca7ee8067d6c5c30695582326ff Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 3 Oct 2025 23:24:42 +0300 Subject: [PATCH 091/114] convert nodes_edit_model.py to V3 schema (#10147) --- comfy_extras/nodes_edit_model.py | 46 ++++++++++++++++++++------------ 1 file changed, 29 insertions(+), 17 deletions(-) diff --git a/comfy_extras/nodes_edit_model.py b/comfy_extras/nodes_edit_model.py index b69f79715..36da66f34 100644 --- a/comfy_extras/nodes_edit_model.py +++ b/comfy_extras/nodes_edit_model.py @@ -1,26 +1,38 @@ import node_helpers +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io -class ReferenceLatent: +class ReferenceLatent(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": {"conditioning": ("CONDITIONING", ), - }, - "optional": {"latent": ("LATENT", ),} - } + def define_schema(cls): + return io.Schema( + node_id="ReferenceLatent", + category="advanced/conditioning/edit_models", + description="This node sets the guiding latent for an edit model. If the model supports it you can chain multiple to set multiple reference images.", + inputs=[ + io.Conditioning.Input("conditioning"), + io.Latent.Input("latent", optional=True), + ], + outputs=[ + io.Conditioning.Output(), + ] + ) - RETURN_TYPES = ("CONDITIONING",) - FUNCTION = "append" - - CATEGORY = "advanced/conditioning/edit_models" - DESCRIPTION = "This node sets the guiding latent for an edit model. If the model supports it you can chain multiple to set multiple reference images." - - def append(self, conditioning, latent=None): + @classmethod + def execute(cls, conditioning, latent=None) -> io.NodeOutput: if latent is not None: conditioning = node_helpers.conditioning_set_values(conditioning, {"reference_latents": [latent["samples"]]}, append=True) - return (conditioning, ) + return io.NodeOutput(conditioning) -NODE_CLASS_MAPPINGS = { - "ReferenceLatent": ReferenceLatent, -} +class EditModelExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + ReferenceLatent, + ] + + +def comfy_entrypoint() -> EditModelExtension: + return EditModelExtension() From 93d859cfaaad150c2a1e5e54c8f14765fa79ecb5 Mon Sep 17 00:00:00 2001 From: Finn-Hecker Date: Fri, 3 Oct 2025 23:32:19 +0200 Subject: [PATCH 092/114] Fix type annotation syntax in MotionEncoder_tc __init__ (#10186) ## Summary Fixed incorrect type hint syntax in `MotionEncoder_tc.__init__()` parameter list. ## Changes - Line 647: Changed `num_heads=int` to `num_heads: int` - This corrects the parameter annotation from a default value assignment to proper type hint syntax ## Details The parameter was using assignment syntax (`=`) instead of type annotation syntax (`:`), which would incorrectly set the default value to the `int` class itself rather than annotating the expected type. --- comfy/ldm/wan/model.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/comfy/ldm/wan/model.py b/comfy/ldm/wan/model.py index 0dc650ced..90c347d3d 100644 --- a/comfy/ldm/wan/model.py +++ b/comfy/ldm/wan/model.py @@ -903,7 +903,7 @@ class MotionEncoder_tc(nn.Module): def __init__(self, in_dim: int, hidden_dim: int, - num_heads=int, + num_heads: int, need_global=True, dtype=None, device=None, From 08726b64fe767f47bf074a05bedd6db45314c4c9 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Fri, 3 Oct 2025 15:22:43 -0700 Subject: [PATCH 093/114] Update amd nightly command in readme. (#10189) --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 8f24a33ee..1224a6176 100644 --- a/README.md +++ b/README.md @@ -211,9 +211,9 @@ AMD users can install rocm and pytorch with pip if you don't have it already ins ```pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/rocm6.4``` -This is the command to install the nightly with ROCm 6.4 which might have some performance improvements: +This is the command to install the nightly with ROCm 7.0 which might have some performance improvements: -```pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/rocm6.4``` +```pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/rocm7.0``` ### Intel GPUs (Windows and Linux) From bbd683098e7d18700f025b2f0a4f6a44a3176602 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Fri, 3 Oct 2025 20:37:43 -0700 Subject: [PATCH 094/114] Add instructions to install nightly AMD pytorch for windows. (#10190) * Add instructions to install nightly AMD pytorch for windows. * Update README.md --- README.md | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-) diff --git a/README.md b/README.md index 1224a6176..4a5a17cda 100644 --- a/README.md +++ b/README.md @@ -206,7 +206,8 @@ Put your SD checkpoints (the huge ckpt/safetensors files) in: models/checkpoints Put your VAE in: models/vae -### AMD GPUs (Linux only) +### AMD GPUs (Linux) + AMD users can install rocm and pytorch with pip if you don't have it already installed, this is the command to install the stable version: ```pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/rocm6.4``` @@ -215,6 +216,23 @@ This is the command to install the nightly with ROCm 7.0 which might have some p ```pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/rocm7.0``` + +### AMD GPUs (Experimental: Windows and Linux), RDNA 3, 3.5 and 4 only. + +These have less hardware support than the builds above but they work on windows. You also need to install the pytorch version specific to your hardware. + +RDNA 3 (RX 7000 series): + +```pip install --pre torch torchvision torchaudio --index-url https://rocm.nightlies.amd.com/v2/gfx110X-dgpu/``` + +RDNA 3.5 (Strix halo/Ryzen AI Max+ 365): + +```pip install --pre torch torchvision torchaudio --index-url https://rocm.nightlies.amd.com/v2/gfx1151/``` + +RDNA 4 (RX 9000 series): + +```pip install --pre torch torchvision torchaudio --index-url https://rocm.nightlies.amd.com/v2/gfx120X-all/``` + ### Intel GPUs (Windows and Linux) (Option 1) Intel Arc GPU users can install native PyTorch with torch.xpu support using pip. More information can be found [here](https://pytorch.org/docs/main/notes/get_start_xpu.html) @@ -270,12 +288,6 @@ You can install ComfyUI in Apple Mac silicon (M1 or M2) with any recent macOS ve > **Note**: Remember to add your models, VAE, LoRAs etc. to the corresponding Comfy folders, as discussed in [ComfyUI manual installation](#manual-install-windows-linux). -#### DirectML (AMD Cards on Windows) - -This is very badly supported and is not recommended. There are some unofficial builds of pytorch ROCm on windows that exist that will give you a much better experience than this. This readme will be updated once official pytorch ROCm builds for windows come out. - -```pip install torch-directml``` Then you can launch ComfyUI with: ```python main.py --directml``` - #### Ascend NPUs For models compatible with Ascend Extension for PyTorch (torch_npu). To get started, ensure your environment meets the prerequisites outlined on the [installation](https://ascend.github.io/docs/sources/ascend/quick_install.html) page. Here's a step-by-step guide tailored to your platform and installation method: From 22f99fb97edaccf450152c8bf7c4068c1d331899 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 4 Oct 2025 22:22:57 +0300 Subject: [PATCH 095/114] fix(api-nodes): enable 2 more pylint rules, removed non needed code (#10192) --- comfy_api_nodes/nodes_gemini.py | 3 +- comfy_api_nodes/nodes_moonvalley.py | 49 ++--------------------------- pyproject.toml | 2 -- 3 files changed, 4 insertions(+), 50 deletions(-) diff --git a/comfy_api_nodes/nodes_gemini.py b/comfy_api_nodes/nodes_gemini.py index 151cb4044..309e9a2d2 100644 --- a/comfy_api_nodes/nodes_gemini.py +++ b/comfy_api_nodes/nodes_gemini.py @@ -39,6 +39,7 @@ from comfy_api_nodes.apinode_utils import ( tensor_to_base64_string, bytesio_to_image_tensor, ) +from comfy_api.util import VideoContainer, VideoCodec GEMINI_BASE_ENDPOINT = "/proxy/vertexai/gemini" @@ -310,7 +311,7 @@ class GeminiNode(ComfyNodeABC): Returns: List of GeminiPart objects containing the encoded video. """ - from comfy_api.util import VideoContainer, VideoCodec + base_64_string = video_to_base64_string( video_input, container_format=VideoContainer.MP4, diff --git a/comfy_api_nodes/nodes_moonvalley.py b/comfy_api_nodes/nodes_moonvalley.py index 08e838fef..6467dd614 100644 --- a/comfy_api_nodes/nodes_moonvalley.py +++ b/comfy_api_nodes/nodes_moonvalley.py @@ -2,11 +2,7 @@ import logging from typing import Any, Callable, Optional, TypeVar import torch from typing_extensions import override -from comfy_api_nodes.util.validation_utils import ( - get_image_dimensions, - validate_image_dimensions, -) - +from comfy_api_nodes.util.validation_utils import validate_image_dimensions from comfy_api_nodes.apis import ( MoonvalleyTextToVideoRequest, @@ -132,47 +128,6 @@ def validate_prompts( return True -def validate_input_media(width, height, with_frame_conditioning, num_frames_in=None): - # inference validation - # T = num_frames - # in all cases, the following must be true: T divisible by 16 and H,W by 8. in addition... - # with image conditioning: H*W must be divisible by 8192 - # without image conditioning: T divisible by 32 - if num_frames_in and not num_frames_in % 16 == 0: - return False, ("The input video total frame count must be divisible by 16!") - - if height % 8 != 0 or width % 8 != 0: - return False, ( - f"Height ({height}) and width ({width}) must be " "divisible by 8" - ) - - if with_frame_conditioning: - if (height * width) % 8192 != 0: - return False, ( - f"Height * width ({height * width}) must be " - "divisible by 8192 for frame conditioning" - ) - else: - if num_frames_in and not num_frames_in % 32 == 0: - return False, ("The input video total frame count must be divisible by 32!") - - -def validate_input_image( - image: torch.Tensor, with_frame_conditioning: bool = False -) -> None: - """ - Validates the input image adheres to the expectations of the API: - - The image resolution should not be less than 300*300px - - The aspect ratio of the image should be between 1:2.5 ~ 2.5:1 - - """ - height, width = get_image_dimensions(image) - validate_input_media(width, height, with_frame_conditioning) - validate_image_dimensions( - image, min_width=300, min_height=300, max_height=MAX_HEIGHT, max_width=MAX_WIDTH - ) - - def validate_video_to_video_input(video: VideoInput) -> VideoInput: """ Validates and processes video input for Moonvalley Video-to-Video generation. @@ -499,7 +454,7 @@ class MoonvalleyImg2VideoNode(comfy_io.ComfyNode): seed: int, steps: int, ) -> comfy_io.NodeOutput: - validate_input_image(image, True) + validate_image_dimensions(image, min_width=300, min_height=300, max_height=MAX_HEIGHT, max_width=MAX_WIDTH) validate_prompts(prompt, negative_prompt, MOONVALLEY_MAREY_MAX_PROMPT_LENGTH) width_height = parse_width_height_from_res(resolution) diff --git a/pyproject.toml b/pyproject.toml index 240919a43..383e7d10a 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -70,7 +70,5 @@ messages_control.disable = [ "invalid-overridden-method", "unused-variable", "pointless-string-statement", - "inconsistent-return-statements", - "import-outside-toplevel", "redefined-outer-name", ] From 2ed74f7ac78d3ff713d0a8583695c31055914b76 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 4 Oct 2025 22:29:09 +0300 Subject: [PATCH 096/114] convert nodes_rodin.py to V3 schema (#10195) --- comfy_api_nodes/nodes_rodin.py | 941 +++++++++++++++++---------------- 1 file changed, 478 insertions(+), 463 deletions(-) diff --git a/comfy_api_nodes/nodes_rodin.py b/comfy_api_nodes/nodes_rodin.py index 633ac46d3..bd758f762 100644 --- a/comfy_api_nodes/nodes_rodin.py +++ b/comfy_api_nodes/nodes_rodin.py @@ -7,14 +7,15 @@ Rodin API docs: https://developer.hyper3d.ai/ from __future__ import annotations from inspect import cleandoc -from comfy.comfy_types.node_typing import IO import folder_paths as comfy_paths import aiohttp import os import asyncio -import io import logging import math +from typing import Optional +from io import BytesIO +from typing_extensions import override from PIL import Image from comfy_api_nodes.apis.rodin_api import ( Rodin3DGenerateRequest, @@ -31,428 +32,436 @@ from comfy_api_nodes.apis.client import ( SynchronousOperation, PollingOperation, ) +from comfy_api.latest import ComfyExtension, io as comfy_io -COMMON_PARAMETERS = { - "Seed": ( - IO.INT, - { - "default":0, - "min":0, - "max":65535, - "display":"number" - } +COMMON_PARAMETERS = [ + comfy_io.Int.Input( + "Seed", + default=0, + min=0, + max=65535, + display_mode=comfy_io.NumberDisplay.number, + optional=True, ), - "Material_Type": ( - IO.COMBO, - { - "options": ["PBR", "Shaded"], - "default": "PBR" - } + comfy_io.Combo.Input("Material_Type", options=["PBR", "Shaded"], default="PBR", optional=True), + comfy_io.Combo.Input( + "Polygon_count", + options=["4K-Quad", "8K-Quad", "18K-Quad", "50K-Quad", "200K-Triangle"], + default="18K-Quad", + optional=True, ), - "Polygon_count": ( - IO.COMBO, - { - "options": ["4K-Quad", "8K-Quad", "18K-Quad", "50K-Quad", "200K-Triangle"], - "default": "18K-Quad" - } +] + + +def get_quality_mode(poly_count): + polycount = poly_count.split("-") + poly = polycount[1] + count = polycount[0] + if poly == "Triangle": + mesh_mode = "Raw" + elif poly == "Quad": + mesh_mode = "Quad" + else: + mesh_mode = "Quad" + + if count == "4K": + quality_override = 4000 + elif count == "8K": + quality_override = 8000 + elif count == "18K": + quality_override = 18000 + elif count == "50K": + quality_override = 50000 + elif count == "2K": + quality_override = 2000 + elif count == "20K": + quality_override = 20000 + elif count == "150K": + quality_override = 150000 + elif count == "500K": + quality_override = 500000 + else: + quality_override = 18000 + + return mesh_mode, quality_override + + +def tensor_to_filelike(tensor, max_pixels: int = 2048*2048): + """ + Converts a PyTorch tensor to a file-like object. + + Args: + - tensor (torch.Tensor): A tensor representing an image of shape (H, W, C) + where C is the number of channels (3 for RGB), H is height, and W is width. + + Returns: + - io.BytesIO: A file-like object containing the image data. + """ + array = tensor.cpu().numpy() + array = (array * 255).astype('uint8') + image = Image.fromarray(array, 'RGB') + + original_width, original_height = image.size + original_pixels = original_width * original_height + if original_pixels > max_pixels: + scale = math.sqrt(max_pixels / original_pixels) + new_width = int(original_width * scale) + new_height = int(original_height * scale) + else: + new_width, new_height = original_width, original_height + + if new_width != original_width or new_height != original_height: + image = image.resize((new_width, new_height), Image.Resampling.LANCZOS) + + img_byte_arr = BytesIO() + image.save(img_byte_arr, format='PNG') # PNG is used for lossless compression + img_byte_arr.seek(0) + return img_byte_arr + + +async def create_generate_task( + images=None, + seed=1, + material="PBR", + quality_override=18000, + tier="Regular", + mesh_mode="Quad", + TAPose = False, + auth_kwargs: Optional[dict[str, str]] = None, +): + if images is None: + raise Exception("Rodin 3D generate requires at least 1 image.") + if len(images) > 5: + raise Exception("Rodin 3D generate requires up to 5 image.") + + path = "/proxy/rodin/api/v2/rodin" + operation = SynchronousOperation( + endpoint=ApiEndpoint( + path=path, + method=HttpMethod.POST, + request_model=Rodin3DGenerateRequest, + response_model=Rodin3DGenerateResponse, + ), + request=Rodin3DGenerateRequest( + seed=seed, + tier=tier, + material=material, + quality_override=quality_override, + mesh_mode=mesh_mode, + TAPose=TAPose, + ), + files=[ + ( + "images", + open(image, "rb") if isinstance(image, str) else tensor_to_filelike(image) + ) + for image in images if image is not None + ], + content_type="multipart/form-data", + auth_kwargs=auth_kwargs, ) -} -def create_task_error(response: Rodin3DGenerateResponse): - """Check if the response has error""" - return hasattr(response, "error") + response = await operation.execute() + + if hasattr(response, "error"): + error_message = f"Rodin3D Create 3D generate Task Failed. Message: {response.message}, error: {response.error}" + logging.error(error_message) + raise Exception(error_message) + + logging.info("[ Rodin3D API - Submit Jobs ] Submit Generate Task Success!") + subscription_key = response.jobs.subscription_key + task_uuid = response.uuid + logging.info(f"[ Rodin3D API - Submit Jobs ] UUID: {task_uuid}") + return task_uuid, subscription_key -class Rodin3DAPI: - """ - Generate 3D Assets using Rodin API - """ - RETURN_TYPES = (IO.STRING,) - RETURN_NAMES = ("3D Model Path",) - CATEGORY = "api node/3d/Rodin" - DESCRIPTION = cleandoc(__doc__ or "") - FUNCTION = "api_call" - API_NODE = True - - def tensor_to_filelike(self, tensor, max_pixels: int = 2048*2048): - """ - Converts a PyTorch tensor to a file-like object. - - Args: - - tensor (torch.Tensor): A tensor representing an image of shape (H, W, C) - where C is the number of channels (3 for RGB), H is height, and W is width. - - Returns: - - io.BytesIO: A file-like object containing the image data. - """ - array = tensor.cpu().numpy() - array = (array * 255).astype('uint8') - image = Image.fromarray(array, 'RGB') - - original_width, original_height = image.size - original_pixels = original_width * original_height - if original_pixels > max_pixels: - scale = math.sqrt(max_pixels / original_pixels) - new_width = int(original_width * scale) - new_height = int(original_height * scale) - else: - new_width, new_height = original_width, original_height - - if new_width != original_width or new_height != original_height: - image = image.resize((new_width, new_height), Image.Resampling.LANCZOS) - - img_byte_arr = io.BytesIO() - image.save(img_byte_arr, format='PNG') # PNG is used for lossless compression - img_byte_arr.seek(0) - return img_byte_arr - - def check_rodin_status(self, response: Rodin3DCheckStatusResponse) -> str: - has_failed = any(job.status == JobStatus.Failed for job in response.jobs) - all_done = all(job.status == JobStatus.Done for job in response.jobs) - status_list = [str(job.status) for job in response.jobs] - logging.info(f"[ Rodin3D API - CheckStatus ] Generate Status: {status_list}") - if has_failed: - logging.error(f"[ Rodin3D API - CheckStatus ] Generate Failed: {status_list}, Please try again.") - raise Exception("[ Rodin3D API ] Generate Failed, Please Try again.") - elif all_done: - return "DONE" - else: - return "Generating" - - async def create_generate_task(self, images=None, seed=1, material="PBR", quality_override=18000, tier="Regular", mesh_mode="Quad", TAPose = False, **kwargs): - if images is None: - raise Exception("Rodin 3D generate requires at least 1 image.") - if len(images) > 5: - raise Exception("Rodin 3D generate requires up to 5 image.") - - path = "/proxy/rodin/api/v2/rodin" - operation = SynchronousOperation( - endpoint=ApiEndpoint( - path=path, - method=HttpMethod.POST, - request_model=Rodin3DGenerateRequest, - response_model=Rodin3DGenerateResponse, - ), - request=Rodin3DGenerateRequest( - seed=seed, - tier=tier, - material=material, - quality_override=quality_override, - mesh_mode=mesh_mode, - TAPose=TAPose, - ), - files=[ - ( - "images", - open(image, "rb") if isinstance(image, str) else self.tensor_to_filelike(image) - ) - for image in images if image is not None - ], - content_type = "multipart/form-data", - auth_kwargs=kwargs, - ) - - response = await operation.execute() - - if create_task_error(response): - error_message = f"Rodin3D Create 3D generate Task Failed. Message: {response.message}, error: {response.error}" - logging.error(error_message) - raise Exception(error_message) - - logging.info("[ Rodin3D API - Submit Jobs ] Submit Generate Task Success!") - subscription_key = response.jobs.subscription_key - task_uuid = response.uuid - logging.info(f"[ Rodin3D API - Submit Jobs ] UUID: {task_uuid}") - return task_uuid, subscription_key - - async def poll_for_task_status(self, subscription_key, **kwargs) -> Rodin3DCheckStatusResponse: - - path = "/proxy/rodin/api/v2/status" - - poll_operation = PollingOperation( - poll_endpoint=ApiEndpoint( - path = path, - method=HttpMethod.POST, - request_model=Rodin3DCheckStatusRequest, - response_model=Rodin3DCheckStatusResponse, - ), - request=Rodin3DCheckStatusRequest( - subscription_key = subscription_key - ), - completed_statuses=["DONE"], - failed_statuses=["FAILED"], - status_extractor=self.check_rodin_status, - poll_interval=3.0, - auth_kwargs=kwargs, - ) - - logging.info("[ Rodin3D API - CheckStatus ] Generate Start!") - - return await poll_operation.execute() - - async def get_rodin_download_list(self, uuid, **kwargs) -> Rodin3DDownloadResponse: - logging.info("[ Rodin3D API - Downloading ] Generate Successfully!") - - path = "/proxy/rodin/api/v2/download" - operation = SynchronousOperation( - endpoint=ApiEndpoint( - path=path, - method=HttpMethod.POST, - request_model=Rodin3DDownloadRequest, - response_model=Rodin3DDownloadResponse, - ), - request=Rodin3DDownloadRequest( - task_uuid=uuid - ), - auth_kwargs=kwargs - ) - - return await operation.execute() - - def get_quality_mode(self, poly_count): - polycount = poly_count.split("-") - poly = polycount[1] - count = polycount[0] - if poly == "Triangle": - mesh_mode = "Raw" - elif poly == "Quad": - mesh_mode = "Quad" - else: - mesh_mode = "Quad" - - if count == "4K": - quality_override = 4000 - elif count == "8K": - quality_override = 8000 - elif count == "18K": - quality_override = 18000 - elif count == "50K": - quality_override = 50000 - elif count == "2K": - quality_override = 2000 - elif count == "20K": - quality_override = 20000 - elif count == "150K": - quality_override = 150000 - elif count == "500K": - quality_override = 500000 - else: - quality_override = 18000 - - return mesh_mode, quality_override - - async def download_files(self, url_list, task_uuid): - save_path = os.path.join(comfy_paths.get_output_directory(), f"Rodin3D_{task_uuid}") - os.makedirs(save_path, exist_ok=True) - model_file_path = None - async with aiohttp.ClientSession() as session: - for i in url_list.list: - url = i.url - file_name = i.name - file_path = os.path.join(save_path, file_name) - if file_path.endswith(".glb"): - model_file_path = file_path - logging.info(f"[ Rodin3D API - download_files ] Downloading file: {file_path}") - max_retries = 5 - for attempt in range(max_retries): - try: - async with session.get(url) as resp: - resp.raise_for_status() - with open(file_path, "wb") as f: - async for chunk in resp.content.iter_chunked(32 * 1024): - f.write(chunk) - break - except Exception as e: - logging.info(f"[ Rodin3D API - download_files ] Error downloading {file_path}:{e}") - if attempt < max_retries - 1: - logging.info("Retrying...") - await asyncio.sleep(2) - else: - logging.info( - "[ Rodin3D API - download_files ] Failed to download %s after %s attempts.", - file_path, - max_retries, - ) - - return model_file_path +def check_rodin_status(response: Rodin3DCheckStatusResponse) -> str: + all_done = all(job.status == JobStatus.Done for job in response.jobs) + status_list = [str(job.status) for job in response.jobs] + logging.info(f"[ Rodin3D API - CheckStatus ] Generate Status: {status_list}") + if any(job.status == JobStatus.Failed for job in response.jobs): + logging.error(f"[ Rodin3D API - CheckStatus ] Generate Failed: {status_list}, Please try again.") + raise Exception("[ Rodin3D API ] Generate Failed, Please Try again.") + if all_done: + return "DONE" + return "Generating" -class Rodin3D_Regular(Rodin3DAPI): +async def poll_for_task_status( + subscription_key, auth_kwargs: Optional[dict[str, str]] = None, +) -> Rodin3DCheckStatusResponse: + poll_operation = PollingOperation( + poll_endpoint=ApiEndpoint( + path="/proxy/rodin/api/v2/status", + method=HttpMethod.POST, + request_model=Rodin3DCheckStatusRequest, + response_model=Rodin3DCheckStatusResponse, + ), + request=Rodin3DCheckStatusRequest(subscription_key=subscription_key), + completed_statuses=["DONE"], + failed_statuses=["FAILED"], + status_extractor=check_rodin_status, + poll_interval=3.0, + auth_kwargs=auth_kwargs, + ) + logging.info("[ Rodin3D API - CheckStatus ] Generate Start!") + return await poll_operation.execute() + + +async def get_rodin_download_list(uuid, auth_kwargs: Optional[dict[str, str]] = None) -> Rodin3DDownloadResponse: + logging.info("[ Rodin3D API - Downloading ] Generate Successfully!") + operation = SynchronousOperation( + endpoint=ApiEndpoint( + path="/proxy/rodin/api/v2/download", + method=HttpMethod.POST, + request_model=Rodin3DDownloadRequest, + response_model=Rodin3DDownloadResponse, + ), + request=Rodin3DDownloadRequest(task_uuid=uuid), + auth_kwargs=auth_kwargs, + ) + return await operation.execute() + + +async def download_files(url_list, task_uuid): + save_path = os.path.join(comfy_paths.get_output_directory(), f"Rodin3D_{task_uuid}") + os.makedirs(save_path, exist_ok=True) + model_file_path = None + async with aiohttp.ClientSession() as session: + for i in url_list.list: + url = i.url + file_name = i.name + file_path = os.path.join(save_path, file_name) + if file_path.endswith(".glb"): + model_file_path = file_path + logging.info(f"[ Rodin3D API - download_files ] Downloading file: {file_path}") + max_retries = 5 + for attempt in range(max_retries): + try: + async with session.get(url) as resp: + resp.raise_for_status() + with open(file_path, "wb") as f: + async for chunk in resp.content.iter_chunked(32 * 1024): + f.write(chunk) + break + except Exception as e: + logging.info(f"[ Rodin3D API - download_files ] Error downloading {file_path}:{e}") + if attempt < max_retries - 1: + logging.info("Retrying...") + await asyncio.sleep(2) + else: + logging.info( + "[ Rodin3D API - download_files ] Failed to download %s after %s attempts.", + file_path, + max_retries, + ) + return model_file_path + + +class Rodin3D_Regular(comfy_io.ComfyNode): + """Generate 3D Assets using Rodin API""" + @classmethod - def INPUT_TYPES(s): - return { - "required": { - "Images": - ( - IO.IMAGE, - { - "forceInput":True, - } - ) - }, - "optional": { - **COMMON_PARAMETERS - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - }, - } + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="Rodin3D_Regular", + display_name="Rodin 3D Generate - Regular Generate", + category="api node/3d/Rodin", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("Images"), + *COMMON_PARAMETERS, + ], + outputs=[comfy_io.String.Output(display_name="3D Model Path")], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + ], + is_api_node=True, + ) - async def api_call( - self, + @classmethod + async def execute( + cls, Images, Seed, Material_Type, Polygon_count, - **kwargs - ): + ) -> comfy_io.NodeOutput: tier = "Regular" num_images = Images.shape[0] m_images = [] for i in range(num_images): m_images.append(Images[i]) - mesh_mode, quality_override = self.get_quality_mode(Polygon_count) - task_uuid, subscription_key = await self.create_generate_task(images=m_images, seed=Seed, material=Material_Type, - quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, - **kwargs) - await self.poll_for_task_status(subscription_key, **kwargs) - download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list, task_uuid) - - return (model,) - - -class Rodin3D_Detail(Rodin3DAPI): - @classmethod - def INPUT_TYPES(s): - return { - "required": { - "Images": - ( - IO.IMAGE, - { - "forceInput":True, - } - ) - }, - "optional": { - **COMMON_PARAMETERS - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - }, + mesh_mode, quality_override = get_quality_mode(Polygon_count) + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, } + task_uuid, subscription_key = await create_generate_task( + images=m_images, + seed=Seed, + material=Material_Type, + quality_override=quality_override, + tier=tier, + mesh_mode=mesh_mode, + auth_kwargs=auth, + ) + await poll_for_task_status(subscription_key, auth_kwargs=auth) + download_list = await get_rodin_download_list(task_uuid, auth_kwargs=auth) + model = await download_files(download_list, task_uuid) - async def api_call( - self, + return comfy_io.NodeOutput(model) + + +class Rodin3D_Detail(comfy_io.ComfyNode): + """Generate 3D Assets using Rodin API""" + + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="Rodin3D_Detail", + display_name="Rodin 3D Generate - Detail Generate", + category="api node/3d/Rodin", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("Images"), + *COMMON_PARAMETERS, + ], + outputs=[comfy_io.String.Output(display_name="3D Model Path")], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, Images, Seed, Material_Type, Polygon_count, - **kwargs - ): + ) -> comfy_io.NodeOutput: tier = "Detail" num_images = Images.shape[0] m_images = [] for i in range(num_images): m_images.append(Images[i]) - mesh_mode, quality_override = self.get_quality_mode(Polygon_count) - task_uuid, subscription_key = await self.create_generate_task(images=m_images, seed=Seed, material=Material_Type, - quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, - **kwargs) - await self.poll_for_task_status(subscription_key, **kwargs) - download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list, task_uuid) - - return (model,) - - -class Rodin3D_Smooth(Rodin3DAPI): - @classmethod - def INPUT_TYPES(s): - return { - "required": { - "Images": - ( - IO.IMAGE, - { - "forceInput":True, - } - ) - }, - "optional": { - **COMMON_PARAMETERS - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - }, + mesh_mode, quality_override = get_quality_mode(Polygon_count) + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, } + task_uuid, subscription_key = await create_generate_task( + images=m_images, + seed=Seed, + material=Material_Type, + quality_override=quality_override, + tier=tier, + mesh_mode=mesh_mode, + auth_kwargs=auth, + ) + await poll_for_task_status(subscription_key, auth_kwargs=auth) + download_list = await get_rodin_download_list(task_uuid, auth_kwargs=auth) + model = await download_files(download_list, task_uuid) - async def api_call( - self, + return comfy_io.NodeOutput(model) + + +class Rodin3D_Smooth(comfy_io.ComfyNode): + """Generate 3D Assets using Rodin API""" + + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="Rodin3D_Smooth", + display_name="Rodin 3D Generate - Smooth Generate", + category="api node/3d/Rodin", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("Images"), + *COMMON_PARAMETERS, + ], + outputs=[comfy_io.String.Output(display_name="3D Model Path")], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, Images, Seed, Material_Type, Polygon_count, - **kwargs - ): + ) -> comfy_io.NodeOutput: tier = "Smooth" num_images = Images.shape[0] m_images = [] for i in range(num_images): m_images.append(Images[i]) - mesh_mode, quality_override = self.get_quality_mode(Polygon_count) - task_uuid, subscription_key = await self.create_generate_task(images=m_images, seed=Seed, material=Material_Type, - quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, - **kwargs) - await self.poll_for_task_status(subscription_key, **kwargs) - download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list, task_uuid) - - return (model,) - - -class Rodin3D_Sketch(Rodin3DAPI): - @classmethod - def INPUT_TYPES(s): - return { - "required": { - "Images": - ( - IO.IMAGE, - { - "forceInput":True, - } - ) - }, - "optional": { - "Seed": - ( - IO.INT, - { - "default":0, - "min":0, - "max":65535, - "display":"number" - } - ) - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - }, + mesh_mode, quality_override = get_quality_mode(Polygon_count) + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, } + task_uuid, subscription_key = await create_generate_task( + images=m_images, + seed=Seed, + material=Material_Type, + quality_override=quality_override, + tier=tier, + mesh_mode=mesh_mode, + auth_kwargs=auth, + ) + await poll_for_task_status(subscription_key, auth_kwargs=auth) + download_list = await get_rodin_download_list(task_uuid, auth_kwargs=auth) + model = await download_files(download_list, task_uuid) - async def api_call( - self, + return comfy_io.NodeOutput(model) + + +class Rodin3D_Sketch(comfy_io.ComfyNode): + """Generate 3D Assets using Rodin API""" + + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="Rodin3D_Sketch", + display_name="Rodin 3D Generate - Sketch Generate", + category="api node/3d/Rodin", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("Images"), + comfy_io.Int.Input( + "Seed", + default=0, + min=0, + max=65535, + display_mode=comfy_io.NumberDisplay.number, + optional=True, + ), + ], + outputs=[comfy_io.String.Output(display_name="3D Model Path")], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, Images, Seed, - **kwargs - ): + ) -> comfy_io.NodeOutput: tier = "Sketch" num_images = Images.shape[0] m_images = [] @@ -461,104 +470,110 @@ class Rodin3D_Sketch(Rodin3DAPI): material_type = "PBR" quality_override = 18000 mesh_mode = "Quad" - task_uuid, subscription_key = await self.create_generate_task( - images=m_images, seed=Seed, material=material_type, quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, **kwargs - ) - await self.poll_for_task_status(subscription_key, **kwargs) - download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list, task_uuid) - - return (model,) - -class Rodin3D_Gen2(Rodin3DAPI): - @classmethod - def INPUT_TYPES(s): - return { - "required": { - "Images": - ( - IO.IMAGE, - { - "forceInput":True, - } - ) - }, - "optional": { - "Seed": ( - IO.INT, - { - "default":0, - "min":0, - "max":65535, - "display":"number" - } - ), - "Material_Type": ( - IO.COMBO, - { - "options": ["PBR", "Shaded"], - "default": "PBR" - } - ), - "Polygon_count": ( - IO.COMBO, - { - "options": ["4K-Quad", "8K-Quad", "18K-Quad", "50K-Quad", "2K-Triangle", "20K-Triangle", "150K-Triangle", "500K-Triangle"], - "default": "500K-Triangle" - } - ), - "TAPose": ( - IO.BOOLEAN, - { - "default": False, - } - ) - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - }, + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, } + task_uuid, subscription_key = await create_generate_task( + images=m_images, + seed=Seed, + material=material_type, + quality_override=quality_override, + tier=tier, + mesh_mode=mesh_mode, + auth_kwargs=auth, + ) + await poll_for_task_status(subscription_key, auth_kwargs=auth) + download_list = await get_rodin_download_list(task_uuid, auth_kwargs=auth) + model = await download_files(download_list, task_uuid) - async def api_call( - self, + return comfy_io.NodeOutput(model) + + +class Rodin3D_Gen2(comfy_io.ComfyNode): + """Generate 3D Assets using Rodin API""" + + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="Rodin3D_Gen2", + display_name="Rodin 3D Generate - Gen-2 Generate", + category="api node/3d/Rodin", + description=cleandoc(cls.__doc__ or ""), + inputs=[ + comfy_io.Image.Input("Images"), + comfy_io.Int.Input( + "Seed", + default=0, + min=0, + max=65535, + display_mode=comfy_io.NumberDisplay.number, + optional=True, + ), + comfy_io.Combo.Input("Material_Type", options=["PBR", "Shaded"], default="PBR", optional=True), + comfy_io.Combo.Input( + "Polygon_count", + options=["4K-Quad", "8K-Quad", "18K-Quad", "50K-Quad", "2K-Triangle", "20K-Triangle", "150K-Triangle", "500K-Triangle"], + default="500K-Triangle", + optional=True, + ), + comfy_io.Boolean.Input("TAPose", default=False), + ], + outputs=[comfy_io.String.Output(display_name="3D Model Path")], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, Images, Seed, Material_Type, Polygon_count, TAPose, - **kwargs - ): + ) -> comfy_io.NodeOutput: tier = "Gen-2" num_images = Images.shape[0] m_images = [] for i in range(num_images): m_images.append(Images[i]) - mesh_mode, quality_override = self.get_quality_mode(Polygon_count) - task_uuid, subscription_key = await self.create_generate_task(images=m_images, seed=Seed, material=Material_Type, - quality_override=quality_override, tier=tier, mesh_mode=mesh_mode, TAPose=TAPose, - **kwargs) - await self.poll_for_task_status(subscription_key, **kwargs) - download_list = await self.get_rodin_download_list(task_uuid, **kwargs) - model = await self.download_files(download_list, task_uuid) + mesh_mode, quality_override = get_quality_mode(Polygon_count) + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } + task_uuid, subscription_key = await create_generate_task( + images=m_images, + seed=Seed, + material=Material_Type, + quality_override=quality_override, + tier=tier, + mesh_mode=mesh_mode, + TAPose=TAPose, + auth_kwargs=auth, + ) + await poll_for_task_status(subscription_key, auth_kwargs=auth) + download_list = await get_rodin_download_list(task_uuid, auth_kwargs=auth) + model = await download_files(download_list, task_uuid) - return (model,) + return comfy_io.NodeOutput(model) -# A dictionary that contains all nodes you want to export with their names -# NOTE: names should be globally unique -NODE_CLASS_MAPPINGS = { - "Rodin3D_Regular": Rodin3D_Regular, - "Rodin3D_Detail": Rodin3D_Detail, - "Rodin3D_Smooth": Rodin3D_Smooth, - "Rodin3D_Sketch": Rodin3D_Sketch, - "Rodin3D_Gen2": Rodin3D_Gen2, -} -# A dictionary that contains the friendly/humanly readable titles for the nodes -NODE_DISPLAY_NAME_MAPPINGS = { - "Rodin3D_Regular": "Rodin 3D Generate - Regular Generate", - "Rodin3D_Detail": "Rodin 3D Generate - Detail Generate", - "Rodin3D_Smooth": "Rodin 3D Generate - Smooth Generate", - "Rodin3D_Sketch": "Rodin 3D Generate - Sketch Generate", - "Rodin3D_Gen2": "Rodin 3D Generate - Gen-2 Generate", -} +class Rodin3DExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: + return [ + Rodin3D_Regular, + Rodin3D_Detail, + Rodin3D_Smooth, + Rodin3D_Sketch, + Rodin3D_Gen2, + ] + + +async def comfy_entrypoint() -> Rodin3DExtension: + return Rodin3DExtension() From b1fa1922df597af759150f4e26ecb276c9753ee4 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sat, 4 Oct 2025 22:33:48 +0300 Subject: [PATCH 097/114] convert nodes_stable3d.py to V3 schema (#10204) --- comfy_extras/nodes_stable3d.py | 149 +++++++++++++++++++-------------- 1 file changed, 86 insertions(+), 63 deletions(-) diff --git a/comfy_extras/nodes_stable3d.py b/comfy_extras/nodes_stable3d.py index be2e34c28..c6d8a683d 100644 --- a/comfy_extras/nodes_stable3d.py +++ b/comfy_extras/nodes_stable3d.py @@ -1,6 +1,8 @@ import torch import nodes import comfy.utils +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io def camera_embeddings(elevation, azimuth): elevation = torch.as_tensor([elevation]) @@ -20,26 +22,31 @@ def camera_embeddings(elevation, azimuth): return embeddings -class StableZero123_Conditioning: +class StableZero123_Conditioning(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "clip_vision": ("CLIP_VISION",), - "init_image": ("IMAGE",), - "vae": ("VAE",), - "width": ("INT", {"default": 256, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 8}), - "height": ("INT", {"default": 256, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 8}), - "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096}), - "elevation": ("FLOAT", {"default": 0.0, "min": -180.0, "max": 180.0, "step": 0.1, "round": False}), - "azimuth": ("FLOAT", {"default": 0.0, "min": -180.0, "max": 180.0, "step": 0.1, "round": False}), - }} - RETURN_TYPES = ("CONDITIONING", "CONDITIONING", "LATENT") - RETURN_NAMES = ("positive", "negative", "latent") + def define_schema(cls): + return io.Schema( + node_id="StableZero123_Conditioning", + category="conditioning/3d_models", + inputs=[ + io.ClipVision.Input("clip_vision"), + io.Image.Input("init_image"), + io.Vae.Input("vae"), + io.Int.Input("width", default=256, min=16, max=nodes.MAX_RESOLUTION, step=8), + io.Int.Input("height", default=256, min=16, max=nodes.MAX_RESOLUTION, step=8), + io.Int.Input("batch_size", default=1, min=1, max=4096), + io.Float.Input("elevation", default=0.0, min=-180.0, max=180.0, step=0.1, round=False), + io.Float.Input("azimuth", default=0.0, min=-180.0, max=180.0, step=0.1, round=False) + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + io.Latent.Output(display_name="latent") + ] + ) - FUNCTION = "encode" - - CATEGORY = "conditioning/3d_models" - - def encode(self, clip_vision, init_image, vae, width, height, batch_size, elevation, azimuth): + @classmethod + def execute(cls, clip_vision, init_image, vae, width, height, batch_size, elevation, azimuth) -> io.NodeOutput: output = clip_vision.encode_image(init_image) pooled = output.image_embeds.unsqueeze(0) pixels = comfy.utils.common_upscale(init_image.movedim(-1,1), width, height, "bilinear", "center").movedim(1,-1) @@ -51,30 +58,35 @@ class StableZero123_Conditioning: positive = [[cond, {"concat_latent_image": t}]] negative = [[torch.zeros_like(pooled), {"concat_latent_image": torch.zeros_like(t)}]] latent = torch.zeros([batch_size, 4, height // 8, width // 8]) - return (positive, negative, {"samples":latent}) + return io.NodeOutput(positive, negative, {"samples":latent}) -class StableZero123_Conditioning_Batched: +class StableZero123_Conditioning_Batched(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "clip_vision": ("CLIP_VISION",), - "init_image": ("IMAGE",), - "vae": ("VAE",), - "width": ("INT", {"default": 256, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 8}), - "height": ("INT", {"default": 256, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 8}), - "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096}), - "elevation": ("FLOAT", {"default": 0.0, "min": -180.0, "max": 180.0, "step": 0.1, "round": False}), - "azimuth": ("FLOAT", {"default": 0.0, "min": -180.0, "max": 180.0, "step": 0.1, "round": False}), - "elevation_batch_increment": ("FLOAT", {"default": 0.0, "min": -180.0, "max": 180.0, "step": 0.1, "round": False}), - "azimuth_batch_increment": ("FLOAT", {"default": 0.0, "min": -180.0, "max": 180.0, "step": 0.1, "round": False}), - }} - RETURN_TYPES = ("CONDITIONING", "CONDITIONING", "LATENT") - RETURN_NAMES = ("positive", "negative", "latent") + def define_schema(cls): + return io.Schema( + node_id="StableZero123_Conditioning_Batched", + category="conditioning/3d_models", + inputs=[ + io.ClipVision.Input("clip_vision"), + io.Image.Input("init_image"), + io.Vae.Input("vae"), + io.Int.Input("width", default=256, min=16, max=nodes.MAX_RESOLUTION, step=8), + io.Int.Input("height", default=256, min=16, max=nodes.MAX_RESOLUTION, step=8), + io.Int.Input("batch_size", default=1, min=1, max=4096), + io.Float.Input("elevation", default=0.0, min=-180.0, max=180.0, step=0.1, round=False), + io.Float.Input("azimuth", default=0.0, min=-180.0, max=180.0, step=0.1, round=False), + io.Float.Input("elevation_batch_increment", default=0.0, min=-180.0, max=180.0, step=0.1, round=False), + io.Float.Input("azimuth_batch_increment", default=0.0, min=-180.0, max=180.0, step=0.1, round=False) + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + io.Latent.Output(display_name="latent") + ] + ) - FUNCTION = "encode" - - CATEGORY = "conditioning/3d_models" - - def encode(self, clip_vision, init_image, vae, width, height, batch_size, elevation, azimuth, elevation_batch_increment, azimuth_batch_increment): + @classmethod + def execute(cls, clip_vision, init_image, vae, width, height, batch_size, elevation, azimuth, elevation_batch_increment, azimuth_batch_increment) -> io.NodeOutput: output = clip_vision.encode_image(init_image) pooled = output.image_embeds.unsqueeze(0) pixels = comfy.utils.common_upscale(init_image.movedim(-1,1), width, height, "bilinear", "center").movedim(1,-1) @@ -93,27 +105,32 @@ class StableZero123_Conditioning_Batched: positive = [[cond, {"concat_latent_image": t}]] negative = [[torch.zeros_like(pooled), {"concat_latent_image": torch.zeros_like(t)}]] latent = torch.zeros([batch_size, 4, height // 8, width // 8]) - return (positive, negative, {"samples":latent, "batch_index": [0] * batch_size}) + return io.NodeOutput(positive, negative, {"samples":latent, "batch_index": [0] * batch_size}) -class SV3D_Conditioning: +class SV3D_Conditioning(io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return {"required": { "clip_vision": ("CLIP_VISION",), - "init_image": ("IMAGE",), - "vae": ("VAE",), - "width": ("INT", {"default": 576, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 8}), - "height": ("INT", {"default": 576, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 8}), - "video_frames": ("INT", {"default": 21, "min": 1, "max": 4096}), - "elevation": ("FLOAT", {"default": 0.0, "min": -90.0, "max": 90.0, "step": 0.1, "round": False}), - }} - RETURN_TYPES = ("CONDITIONING", "CONDITIONING", "LATENT") - RETURN_NAMES = ("positive", "negative", "latent") + def define_schema(cls): + return io.Schema( + node_id="SV3D_Conditioning", + category="conditioning/3d_models", + inputs=[ + io.ClipVision.Input("clip_vision"), + io.Image.Input("init_image"), + io.Vae.Input("vae"), + io.Int.Input("width", default=576, min=16, max=nodes.MAX_RESOLUTION, step=8), + io.Int.Input("height", default=576, min=16, max=nodes.MAX_RESOLUTION, step=8), + io.Int.Input("video_frames", default=21, min=1, max=4096), + io.Float.Input("elevation", default=0.0, min=-90.0, max=90.0, step=0.1, round=False) + ], + outputs=[ + io.Conditioning.Output(display_name="positive"), + io.Conditioning.Output(display_name="negative"), + io.Latent.Output(display_name="latent") + ] + ) - FUNCTION = "encode" - - CATEGORY = "conditioning/3d_models" - - def encode(self, clip_vision, init_image, vae, width, height, video_frames, elevation): + @classmethod + def execute(cls, clip_vision, init_image, vae, width, height, video_frames, elevation) -> io.NodeOutput: output = clip_vision.encode_image(init_image) pooled = output.image_embeds.unsqueeze(0) pixels = comfy.utils.common_upscale(init_image.movedim(-1,1), width, height, "bilinear", "center").movedim(1,-1) @@ -133,11 +150,17 @@ class SV3D_Conditioning: positive = [[pooled, {"concat_latent_image": t, "elevation": elevations, "azimuth": azimuths}]] negative = [[torch.zeros_like(pooled), {"concat_latent_image": torch.zeros_like(t), "elevation": elevations, "azimuth": azimuths}]] latent = torch.zeros([video_frames, 4, height // 8, width // 8]) - return (positive, negative, {"samples":latent}) + return io.NodeOutput(positive, negative, {"samples":latent}) -NODE_CLASS_MAPPINGS = { - "StableZero123_Conditioning": StableZero123_Conditioning, - "StableZero123_Conditioning_Batched": StableZero123_Conditioning_Batched, - "SV3D_Conditioning": SV3D_Conditioning, -} +class Stable3DExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[io.ComfyNode]]: + return [ + StableZero123_Conditioning, + StableZero123_Conditioning_Batched, + SV3D_Conditioning, + ] + +async def comfy_entrypoint() -> Stable3DExtension: + return Stable3DExtension() From caf07331ff1b20f4104b9693ed244d6e22f80b5a Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sat, 4 Oct 2025 19:05:05 -0700 Subject: [PATCH 098/114] Remove soundfile dependency. No more torchaudio load or save. (#10210) --- comfy_extras/nodes_audio.py | 2 +- requirements.txt | 1 - 2 files changed, 1 insertion(+), 2 deletions(-) diff --git a/comfy_extras/nodes_audio.py b/comfy_extras/nodes_audio.py index 51c8b9dd9..1c868fcba 100644 --- a/comfy_extras/nodes_audio.py +++ b/comfy_extras/nodes_audio.py @@ -360,7 +360,7 @@ class RecordAudio: def load(self, audio): audio_path = folder_paths.get_annotated_filepath(audio) - waveform, sample_rate = torchaudio.load(audio_path) + waveform, sample_rate = load(audio_path) audio = {"waveform": waveform.unsqueeze(0), "sample_rate": sample_rate} return (audio, ) diff --git a/requirements.txt b/requirements.txt index 588c5dcf0..6c28f9478 100644 --- a/requirements.txt +++ b/requirements.txt @@ -25,6 +25,5 @@ av>=14.2.0 #non essential dependencies: kornia>=0.7.1 spandrel -soundfile pydantic~=2.0 pydantic-settings~=2.0 From 187f43696dd58f252075d2e3c6873706eb6b5fa1 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Sun, 5 Oct 2025 09:34:18 +0300 Subject: [PATCH 099/114] fix(api-nodes): disable "std" mode for Kling2.5-turbo (#10212) --- comfy_api_nodes/nodes_kling.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/comfy_api_nodes/nodes_kling.py b/comfy_api_nodes/nodes_kling.py index d8646f106..44fccc0c7 100644 --- a/comfy_api_nodes/nodes_kling.py +++ b/comfy_api_nodes/nodes_kling.py @@ -712,6 +712,9 @@ class KlingImage2VideoNode(KlingNodeBase): # Camera control type for image 2 video is always `simple` camera_control.type = KlingCameraControlType.simple + if mode == "std" and model_name == KlingVideoGenModelName.kling_v2_5_turbo.value: + mode = "pro" # October 5: currently "std" mode is not supported for this model + initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_IMAGE_TO_VIDEO, From 195e0b063950f585fe584c5ce7b0b689f8d20ff8 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sun, 5 Oct 2025 12:41:19 -0700 Subject: [PATCH 100/114] Remove useless code. (#10223) --- comfy/ldm/ace/vae/music_dcae_pipeline.py | 11 ----------- 1 file changed, 11 deletions(-) diff --git a/comfy/ldm/ace/vae/music_dcae_pipeline.py b/comfy/ldm/ace/vae/music_dcae_pipeline.py index af81280eb..3c8830c17 100644 --- a/comfy/ldm/ace/vae/music_dcae_pipeline.py +++ b/comfy/ldm/ace/vae/music_dcae_pipeline.py @@ -23,8 +23,6 @@ class MusicDCAE(torch.nn.Module): else: self.source_sample_rate = source_sample_rate - # self.resampler = torchaudio.transforms.Resample(source_sample_rate, 44100) - self.transform = transforms.Compose([ transforms.Normalize(0.5, 0.5), ]) @@ -37,10 +35,6 @@ class MusicDCAE(torch.nn.Module): self.scale_factor = 0.1786 self.shift_factor = -1.9091 - def load_audio(self, audio_path): - audio, sr = torchaudio.load(audio_path) - return audio, sr - def forward_mel(self, audios): mels = [] for i in range(len(audios)): @@ -73,10 +67,8 @@ class MusicDCAE(torch.nn.Module): latent = self.dcae.encoder(mel.unsqueeze(0)) latents.append(latent) latents = torch.cat(latents, dim=0) - # latent_lengths = (audio_lengths / sr * 44100 / 512 / self.time_dimention_multiple).long() latents = (latents - self.shift_factor) * self.scale_factor return latents - # return latents, latent_lengths @torch.no_grad() def decode(self, latents, audio_lengths=None, sr=None): @@ -91,9 +83,7 @@ class MusicDCAE(torch.nn.Module): wav = self.vocoder.decode(mels[0]).squeeze(1) if sr is not None: - # resampler = torchaudio.transforms.Resample(44100, sr).to(latents.device).to(latents.dtype) wav = torchaudio.functional.resample(wav, 44100, sr) - # wav = resampler(wav) else: sr = 44100 pred_wavs.append(wav) @@ -101,7 +91,6 @@ class MusicDCAE(torch.nn.Module): if audio_lengths is not None: pred_wavs = [wav[:, :length].cpu() for wav, length in zip(pred_wavs, audio_lengths)] return torch.stack(pred_wavs) - # return sr, pred_wavs def forward(self, audios, audio_lengths=None, sr=None): latents, latent_lengths = self.encode(audios=audios, audio_lengths=audio_lengths, sr=sr) From 7326e46deeab97219cad32d0624991f9ffea4fe5 Mon Sep 17 00:00:00 2001 From: ComfyUI Wiki Date: Tue, 7 Oct 2025 01:57:00 +0800 Subject: [PATCH 101/114] Update template to 0.1.93 (#10235) * Update template to 0.1.92 * Update template to 0.1.93 --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 6c28f9478..db0486960 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,5 +1,5 @@ comfyui-frontend-package==1.27.7 -comfyui-workflow-templates==0.1.91 +comfyui-workflow-templates==0.1.93 comfyui-embedded-docs==0.2.6 torch torchsde From 6bd3f8eb9ff2d7c74e8ca75ad1f854a6b256b714 Mon Sep 17 00:00:00 2001 From: comfyanonymous Date: Mon, 6 Oct 2025 14:49:04 -0400 Subject: [PATCH 102/114] ComfyUI version 0.3.63 --- comfyui_version.py | 2 +- pyproject.toml | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/comfyui_version.py b/comfyui_version.py index ac76fbe35..c3257d4bf 100644 --- a/comfyui_version.py +++ b/comfyui_version.py @@ -1,3 +1,3 @@ # This file is automatically generated by the build process when version is # updated in pyproject.toml. -__version__ = "0.3.62" +__version__ = "0.3.63" diff --git a/pyproject.toml b/pyproject.toml index 383e7d10a..a9e3de0c6 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -1,6 +1,6 @@ [project] name = "ComfyUI" -version = "0.3.62" +version = "0.3.63" readme = "README.md" license = { file = "LICENSE" } requires-python = ">=3.9" From 6ae35158013e50698e680344ab1f54de0d59fef0 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Tue, 7 Oct 2025 02:05:57 +0300 Subject: [PATCH 103/114] fix(api-nodes): enable more pylint rules (#10213) --- comfy_api_nodes/apinode_utils.py | 2 +- comfy_api_nodes/nodes_moonvalley.py | 3 +-- comfy_api_nodes/nodes_recraft.py | 8 ++++---- pyproject.toml | 6 +----- 4 files changed, 7 insertions(+), 12 deletions(-) diff --git a/comfy_api_nodes/apinode_utils.py b/comfy_api_nodes/apinode_utils.py index 37438f835..5ac3b92aa 100644 --- a/comfy_api_nodes/apinode_utils.py +++ b/comfy_api_nodes/apinode_utils.py @@ -152,7 +152,7 @@ def validate_aspect_ratio( raise TypeError( f"Aspect ratio cannot reduce to any less than {minimum_ratio_str} ({minimum_ratio}), but was {aspect_ratio} ({calculated_ratio})." ) - elif calculated_ratio > maximum_ratio: + if calculated_ratio > maximum_ratio: raise TypeError( f"Aspect ratio cannot reduce to any greater than {maximum_ratio_str} ({maximum_ratio}), but was {aspect_ratio} ({calculated_ratio})." ) diff --git a/comfy_api_nodes/nodes_moonvalley.py b/comfy_api_nodes/nodes_moonvalley.py index 6467dd614..55471a69d 100644 --- a/comfy_api_nodes/nodes_moonvalley.py +++ b/comfy_api_nodes/nodes_moonvalley.py @@ -473,7 +473,7 @@ class MoonvalleyImg2VideoNode(comfy_io.ComfyNode): height=width_height["height"], use_negative_prompts=True, ) - """Upload image to comfy backend to have a URL available for further processing""" + # Get MIME type from tensor - assuming PNG format for image tensors mime_type = "image/png" @@ -591,7 +591,6 @@ class MoonvalleyVideo2VideoNode(comfy_io.ComfyNode): validated_video = validate_video_to_video_input(video) video_url = await upload_video_to_comfyapi(validated_video, auth_kwargs=auth) - """Validate prompts and inference input""" validate_prompts(prompt, negative_prompt) # Only include motion_intensity for Motion Transfer diff --git a/comfy_api_nodes/nodes_recraft.py b/comfy_api_nodes/nodes_recraft.py index a006104b7..0bbb551b8 100644 --- a/comfy_api_nodes/nodes_recraft.py +++ b/comfy_api_nodes/nodes_recraft.py @@ -107,7 +107,7 @@ def recraft_multipart_parser(data, parent_key=None, formatter: callable=None, co # if list already exists exists, just extend list with data for check_list in lists_to_check: for conv_tuple in check_list: - if conv_tuple[0] == parent_key and type(conv_tuple[1]) is list: + if conv_tuple[0] == parent_key and isinstance(conv_tuple[1], list): conv_tuple[1].append(formatter(data)) return True return False @@ -119,7 +119,7 @@ def recraft_multipart_parser(data, parent_key=None, formatter: callable=None, co if formatter is None: formatter = lambda v: v # Multipart representation of value - if type(data) is not dict: + if not isinstance(data, dict): # if list already exists exists, just extend list with data added = handle_converted_lists(data, parent_key, converted_to_check) if added: @@ -136,9 +136,9 @@ def recraft_multipart_parser(data, parent_key=None, formatter: callable=None, co for key, value in data.items(): current_key = key if parent_key is None else f"{parent_key}[{key}]" - if type(value) is dict: + if isinstance(value, dict): converted.extend(recraft_multipart_parser(value, current_key, formatter, next_check).items()) - elif type(value) is list: + elif isinstance(value, list): for ind, list_value in enumerate(value): iter_key = f"{current_key}[]" converted.extend(recraft_multipart_parser(list_value, iter_key, formatter, next_check, is_list=True).items()) diff --git a/pyproject.toml b/pyproject.toml index a9e3de0c6..abd1a5f5c 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -57,18 +57,14 @@ messages_control.disable = [ "redefined-builtin", "unnecessary-lambda", "dangerous-default-value", + "invalid-overridden-method", # next warnings should be fixed in future "bad-classmethod-argument", # Class method should have 'cls' as first argument "wrong-import-order", # Standard imports should be placed before third party imports "logging-fstring-interpolation", # Use lazy % formatting in logging functions "ungrouped-imports", "unnecessary-pass", - "unidiomatic-typecheck", "unnecessary-lambda-assignment", "no-else-return", - "no-else-raise", - "invalid-overridden-method", "unused-variable", - "pointless-string-statement", - "redefined-outer-name", ] From a49007a7b07abfdcb10bc10c23514c48935ea914 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Tue, 7 Oct 2025 02:13:43 +0300 Subject: [PATCH 104/114] fix(api-nodes): allow negative_prompt PixVerse to be multiline (#10196) --- comfy_api_nodes/nodes_pixverse.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/comfy_api_nodes/nodes_pixverse.py b/comfy_api_nodes/nodes_pixverse.py index eb98e9653..2c91bbc65 100644 --- a/comfy_api_nodes/nodes_pixverse.py +++ b/comfy_api_nodes/nodes_pixverse.py @@ -146,7 +146,7 @@ class PixverseTextToVideoNode(comfy_io.ComfyNode): comfy_io.String.Input( "negative_prompt", default="", - force_input=True, + multiline=True, tooltip="An optional text description of undesired elements on an image.", optional=True, ), @@ -284,7 +284,7 @@ class PixverseImageToVideoNode(comfy_io.ComfyNode): comfy_io.String.Input( "negative_prompt", default="", - force_input=True, + multiline=True, tooltip="An optional text description of undesired elements on an image.", optional=True, ), @@ -425,7 +425,7 @@ class PixverseTransitionVideoNode(comfy_io.ComfyNode): comfy_io.String.Input( "negative_prompt", default="", - force_input=True, + multiline=True, tooltip="An optional text description of undesired elements on an image.", optional=True, ), From e77e0a8f8fdcdc53deb8207e0d5b16ca56824a4b Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Tue, 7 Oct 2025 02:20:26 +0300 Subject: [PATCH 105/114] convert nodes_pika.py to V3 schema (#10216) --- comfy_api_nodes/nodes_pika.py | 779 ++++++++++++++++------------------ 1 file changed, 373 insertions(+), 406 deletions(-) diff --git a/comfy_api_nodes/nodes_pika.py b/comfy_api_nodes/nodes_pika.py index a8dc43cb3..0a9f04cc2 100644 --- a/comfy_api_nodes/nodes_pika.py +++ b/comfy_api_nodes/nodes_pika.py @@ -5,14 +5,16 @@ Pika API docs: https://pika-827374fb.mintlify.app/api-reference """ from __future__ import annotations -import io +from io import BytesIO import logging from typing import Optional, TypeVar +from enum import Enum import numpy as np import torch -from comfy.comfy_types.node_typing import IO, ComfyNodeABC, InputTypeOptions +from typing_extensions import override +from comfy_api.latest import ComfyExtension, io as comfy_io from comfy_api.input_impl import VideoFromFile from comfy_api.input_impl.video_types import VideoCodec, VideoContainer, VideoInput from comfy_api_nodes.apinode_utils import ( @@ -20,7 +22,6 @@ from comfy_api_nodes.apinode_utils import ( tensor_to_bytesio, ) from comfy_api_nodes.apis import ( - IngredientsMode, PikaBodyGenerate22C2vGenerate22PikascenesPost, PikaBodyGenerate22I2vGenerate22I2vPost, PikaBodyGenerate22KeyframeGenerate22PikaframesPost, @@ -28,10 +29,7 @@ from comfy_api_nodes.apis import ( PikaBodyGeneratePikadditionsGeneratePikadditionsPost, PikaBodyGeneratePikaffectsGeneratePikaffectsPost, PikaBodyGeneratePikaswapsGeneratePikaswapsPost, - PikaDurationEnum, - Pikaffect, PikaGenerateResponse, - PikaResolutionEnum, PikaVideoResponse, ) from comfy_api_nodes.apis.client import ( @@ -41,7 +39,6 @@ from comfy_api_nodes.apis.client import ( PollingOperation, SynchronousOperation, ) -from comfy_api_nodes.mapper_utils import model_field_to_node_input R = TypeVar("R") @@ -58,6 +55,35 @@ PATH_PIKASCENES = f"/proxy/pika/generate/{PIKA_API_VERSION}/pikascenes" PATH_VIDEO_GET = "/proxy/pika/videos" +class PikaDurationEnum(int, Enum): + integer_5 = 5 + integer_10 = 10 + + +class PikaResolutionEnum(str, Enum): + field_1080p = "1080p" + field_720p = "720p" + + +class Pikaffect(str, Enum): + Cake_ify = "Cake-ify" + Crumble = "Crumble" + Crush = "Crush" + Decapitate = "Decapitate" + Deflate = "Deflate" + Dissolve = "Dissolve" + Explode = "Explode" + Eye_pop = "Eye-pop" + Inflate = "Inflate" + Levitate = "Levitate" + Melt = "Melt" + Peel = "Peel" + Poke = "Poke" + Squish = "Squish" + Ta_da = "Ta-da" + Tear = "Tear" + + class PikaApiError(Exception): """Exception for Pika API errors.""" @@ -74,155 +100,121 @@ def is_valid_initial_response(response: PikaGenerateResponse) -> bool: return hasattr(response, "video_id") and response.video_id is not None -class PikaNodeBase(ComfyNodeABC): - """Base class for Pika nodes.""" +async def poll_for_task_status( + task_id: str, + auth_kwargs: Optional[dict[str, str]] = None, + node_id: Optional[str] = None, +) -> PikaGenerateResponse: + polling_operation = PollingOperation( + poll_endpoint=ApiEndpoint( + path=f"{PATH_VIDEO_GET}/{task_id}", + method=HttpMethod.GET, + request_model=EmptyRequest, + response_model=PikaVideoResponse, + ), + completed_statuses=[ + "finished", + ], + failed_statuses=["failed", "cancelled"], + status_extractor=lambda response: ( + response.status.value if response.status else None + ), + progress_extractor=lambda response: ( + response.progress if hasattr(response, "progress") else None + ), + auth_kwargs=auth_kwargs, + result_url_extractor=lambda response: ( + response.url if hasattr(response, "url") else None + ), + node_id=node_id, + estimated_duration=60 + ) + return await polling_operation.execute() - @classmethod - def get_base_inputs_types( - cls, request_model - ) -> dict[str, tuple[IO, InputTypeOptions]]: - """Get the base required inputs types common to all Pika nodes.""" - return { - "prompt_text": model_field_to_node_input( - IO.STRING, - request_model, - "promptText", - multiline=True, - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - request_model, - "negativePrompt", - multiline=True, - ), - "seed": model_field_to_node_input( - IO.INT, - request_model, - "seed", - min=0, - max=0xFFFFFFFF, - control_after_generate=True, - ), - "resolution": model_field_to_node_input( - IO.COMBO, - request_model, - "resolution", - enum_type=PikaResolutionEnum, - ), - "duration": model_field_to_node_input( - IO.COMBO, - request_model, - "duration", - enum_type=PikaDurationEnum, - ), - } - CATEGORY = "api node/video/Pika" - API_NODE = True - FUNCTION = "api_call" - RETURN_TYPES = ("VIDEO",) +async def execute_task( + initial_operation: SynchronousOperation[R, PikaGenerateResponse], + auth_kwargs: Optional[dict[str, str]] = None, + node_id: Optional[str] = None, +) -> tuple[VideoFromFile]: + """Executes the initial operation then polls for the task status until it is completed. - async def poll_for_task_status( - self, - task_id: str, - auth_kwargs: Optional[dict[str, str]] = None, - node_id: Optional[str] = None, - ) -> PikaGenerateResponse: - polling_operation = PollingOperation( - poll_endpoint=ApiEndpoint( - path=f"{PATH_VIDEO_GET}/{task_id}", - method=HttpMethod.GET, - request_model=EmptyRequest, - response_model=PikaVideoResponse, - ), - completed_statuses=[ - "finished", - ], - failed_statuses=["failed", "cancelled"], - status_extractor=lambda response: ( - response.status.value if response.status else None - ), - progress_extractor=lambda response: ( - response.progress if hasattr(response, "progress") else None - ), - auth_kwargs=auth_kwargs, - result_url_extractor=lambda response: ( - response.url if hasattr(response, "url") else None - ), - node_id=node_id, - estimated_duration=60 + Args: + initial_operation: The initial operation to execute. + auth_kwargs: The authentication token(s) to use for the API call. + + Returns: + A tuple containing the video file as a VIDEO output. + """ + initial_response = await initial_operation.execute() + if not is_valid_initial_response(initial_response): + error_msg = f"Pika initial request failed. Code: {initial_response.code}, Message: {initial_response.message}, Data: {initial_response.data}" + logging.error(error_msg) + raise PikaApiError(error_msg) + + task_id = initial_response.video_id + final_response = await poll_for_task_status(task_id, auth_kwargs, node_id=node_id) + if not is_valid_video_response(final_response): + error_msg = ( + f"Pika task {task_id} succeeded but no video data found in response." ) - return await polling_operation.execute() + logging.error(error_msg) + raise PikaApiError(error_msg) - async def execute_task( - self, - initial_operation: SynchronousOperation[R, PikaGenerateResponse], - auth_kwargs: Optional[dict[str, str]] = None, - node_id: Optional[str] = None, - ) -> tuple[VideoFromFile]: - """Executes the initial operation then polls for the task status until it is completed. + video_url = str(final_response.url) + logging.info("Pika task %s succeeded. Video URL: %s", task_id, video_url) - Args: - initial_operation: The initial operation to execute. - auth_kwargs: The authentication token(s) to use for the API call. - - Returns: - A tuple containing the video file as a VIDEO output. - """ - initial_response = await initial_operation.execute() - if not is_valid_initial_response(initial_response): - error_msg = f"Pika initial request failed. Code: {initial_response.code}, Message: {initial_response.message}, Data: {initial_response.data}" - logging.error(error_msg) - raise PikaApiError(error_msg) - - task_id = initial_response.video_id - final_response = await self.poll_for_task_status(task_id, auth_kwargs) - if not is_valid_video_response(final_response): - error_msg = ( - f"Pika task {task_id} succeeded but no video data found in response." - ) - logging.error(error_msg) - raise PikaApiError(error_msg) - - video_url = str(final_response.url) - logging.info("Pika task %s succeeded. Video URL: %s", task_id, video_url) - - return (await download_url_to_video_output(video_url),) + return (await download_url_to_video_output(video_url),) -class PikaImageToVideoV2_2(PikaNodeBase): +def get_base_inputs_types() -> list[comfy_io.Input]: + """Get the base required inputs types common to all Pika nodes.""" + return [ + comfy_io.String.Input("prompt_text", multiline=True), + comfy_io.String.Input("negative_prompt", multiline=True), + comfy_io.Int.Input("seed", min=0, max=0xFFFFFFFF, control_after_generate=True), + comfy_io.Combo.Input( + "resolution", options=[resolution.value for resolution in PikaResolutionEnum], default="1080p" + ), + comfy_io.Combo.Input( + "duration", options=[duration.value for duration in PikaDurationEnum], default=5 + ), + ] + + +class PikaImageToVideoV2_2(comfy_io.ComfyNode): """Pika 2.2 Image to Video Node.""" @classmethod - def INPUT_TYPES(cls): - return { - "required": { - "image": ( - IO.IMAGE, - {"tooltip": "The image to convert to video"}, - ), - **cls.get_base_inputs_types(PikaBodyGenerate22I2vGenerate22I2vPost), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="PikaImageToVideoNode2_2", + display_name="Pika Image to Video", + description="Sends an image and prompt to the Pika API v2.2 to generate a video.", + category="api node/video/Pika", + inputs=[ + comfy_io.Image.Input("image", tooltip="The image to convert to video"), + *get_base_inputs_types(), + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Sends an image and prompt to the Pika API v2.2 to generate a video." - - async def api_call( - self, + @classmethod + async def execute( + cls, image: torch.Tensor, prompt_text: str, negative_prompt: str, seed: int, resolution: str, duration: int, - unique_id: str, - **kwargs, - ) -> tuple[VideoFromFile]: + ) -> comfy_io.NodeOutput: # Convert image to BytesIO image_bytes_io = tensor_to_bytesio(image) image_bytes_io.seek(0) @@ -237,7 +229,10 @@ class PikaImageToVideoV2_2(PikaNodeBase): resolution=resolution, duration=duration, ) - + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_IMAGE_TO_VIDEO, @@ -248,50 +243,55 @@ class PikaImageToVideoV2_2(PikaNodeBase): request=pika_request_data, files=pika_files, content_type="multipart/form-data", - auth_kwargs=kwargs, + auth_kwargs=auth, ) - - return await self.execute_task(initial_operation, auth_kwargs=kwargs, node_id=unique_id) + return await execute_task(initial_operation, auth_kwargs=auth, node_id=cls.hidden.unique_id) -class PikaTextToVideoNodeV2_2(PikaNodeBase): +class PikaTextToVideoNodeV2_2(comfy_io.ComfyNode): """Pika Text2Video v2.2 Node.""" @classmethod - def INPUT_TYPES(cls): - return { - "required": { - **cls.get_base_inputs_types(PikaBodyGenerate22T2vGenerate22T2vPost), - "aspect_ratio": model_field_to_node_input( - IO.FLOAT, - PikaBodyGenerate22T2vGenerate22T2vPost, - "aspectRatio", + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="PikaTextToVideoNode2_2", + display_name="Pika Text to Video", + description="Sends a text prompt to the Pika API v2.2 to generate a video.", + category="api node/video/Pika", + inputs=[ + *get_base_inputs_types(), + comfy_io.Float.Input( + "aspect_ratio", step=0.001, min=0.4, max=2.5, default=1.7777777777777777, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + tooltip="Aspect ratio (width / height)", + ) + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Sends a text prompt to the Pika API v2.2 to generate a video." - - async def api_call( - self, + @classmethod + async def execute( + cls, prompt_text: str, negative_prompt: str, seed: int, resolution: str, duration: int, aspect_ratio: float, - unique_id: str, - **kwargs, - ) -> tuple[VideoFromFile]: + ) -> comfy_io.NodeOutput: + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_TEXT_TO_VIDEO, @@ -307,62 +307,75 @@ class PikaTextToVideoNodeV2_2(PikaNodeBase): duration=duration, aspectRatio=aspect_ratio, ), - auth_kwargs=kwargs, + auth_kwargs=auth, content_type="application/x-www-form-urlencoded", ) - - return await self.execute_task(initial_operation, auth_kwargs=kwargs, node_id=unique_id) + return await execute_task(initial_operation, auth_kwargs=auth, node_id=cls.hidden.unique_id) -class PikaScenesV2_2(PikaNodeBase): +class PikaScenesV2_2(comfy_io.ComfyNode): """PikaScenes v2.2 Node.""" @classmethod - def INPUT_TYPES(cls): - image_ingredient_input = ( - IO.IMAGE, - {"tooltip": "Image that will be used as ingredient to create a video."}, - ) - return { - "required": { - **cls.get_base_inputs_types( - PikaBodyGenerate22C2vGenerate22PikascenesPost, - ), - "ingredients_mode": model_field_to_node_input( - IO.COMBO, - PikaBodyGenerate22C2vGenerate22PikascenesPost, - "ingredientsMode", - enum_type=IngredientsMode, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="PikaScenesV2_2", + display_name="Pika Scenes (Video Image Composition)", + description="Combine your images to create a video with the objects in them. Upload multiple images as ingredients and generate a high-quality video that incorporates all of them.", + category="api node/video/Pika", + inputs=[ + *get_base_inputs_types(), + comfy_io.Combo.Input( + "ingredients_mode", + options=["creative", "precise"], default="creative", ), - "aspect_ratio": model_field_to_node_input( - IO.FLOAT, - PikaBodyGenerate22C2vGenerate22PikascenesPost, - "aspectRatio", + comfy_io.Float.Input( + "aspect_ratio", step=0.001, min=0.4, max=2.5, default=1.7777777777777777, + tooltip="Aspect ratio (width / height)", ), - }, - "optional": { - "image_ingredient_1": image_ingredient_input, - "image_ingredient_2": image_ingredient_input, - "image_ingredient_3": image_ingredient_input, - "image_ingredient_4": image_ingredient_input, - "image_ingredient_5": image_ingredient_input, - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + comfy_io.Image.Input( + "image_ingredient_1", + optional=True, + tooltip="Image that will be used as ingredient to create a video.", + ), + comfy_io.Image.Input( + "image_ingredient_2", + optional=True, + tooltip="Image that will be used as ingredient to create a video.", + ), + comfy_io.Image.Input( + "image_ingredient_3", + optional=True, + tooltip="Image that will be used as ingredient to create a video.", + ), + comfy_io.Image.Input( + "image_ingredient_4", + optional=True, + tooltip="Image that will be used as ingredient to create a video.", + ), + comfy_io.Image.Input( + "image_ingredient_5", + optional=True, + tooltip="Image that will be used as ingredient to create a video.", + ), + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Combine your images to create a video with the objects in them. Upload multiple images as ingredients and generate a high-quality video that incorporates all of them." - - async def api_call( - self, + @classmethod + async def execute( + cls, prompt_text: str, negative_prompt: str, seed: int, @@ -370,14 +383,12 @@ class PikaScenesV2_2(PikaNodeBase): duration: int, ingredients_mode: str, aspect_ratio: float, - unique_id: str, image_ingredient_1: Optional[torch.Tensor] = None, image_ingredient_2: Optional[torch.Tensor] = None, image_ingredient_3: Optional[torch.Tensor] = None, image_ingredient_4: Optional[torch.Tensor] = None, image_ingredient_5: Optional[torch.Tensor] = None, - **kwargs, - ) -> tuple[VideoFromFile]: + ) -> comfy_io.NodeOutput: # Convert all passed images to BytesIO all_image_bytes_io = [] for image in [ @@ -406,7 +417,10 @@ class PikaScenesV2_2(PikaNodeBase): duration=duration, aspectRatio=aspect_ratio, ) - + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_PIKASCENES, @@ -417,63 +431,54 @@ class PikaScenesV2_2(PikaNodeBase): request=pika_request_data, files=pika_files, content_type="multipart/form-data", - auth_kwargs=kwargs, + auth_kwargs=auth, ) - return await self.execute_task(initial_operation, auth_kwargs=kwargs, node_id=unique_id) + return await execute_task(initial_operation, auth_kwargs=auth, node_id=cls.hidden.unique_id) -class PikAdditionsNode(PikaNodeBase): +class PikAdditionsNode(comfy_io.ComfyNode): """Pika Pikadditions Node. Add an image into a video.""" @classmethod - def INPUT_TYPES(cls): - return { - "required": { - "video": (IO.VIDEO, {"tooltip": "The video to add an image to."}), - "image": (IO.IMAGE, {"tooltip": "The image to add to the video."}), - "prompt_text": model_field_to_node_input( - IO.STRING, - PikaBodyGeneratePikadditionsGeneratePikadditionsPost, - "promptText", - multiline=True, - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - PikaBodyGeneratePikadditionsGeneratePikadditionsPost, - "negativePrompt", - multiline=True, - ), - "seed": model_field_to_node_input( - IO.INT, - PikaBodyGeneratePikadditionsGeneratePikadditionsPost, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="Pikadditions", + display_name="Pikadditions (Video Object Insertion)", + description="Add any object or image into your video. Upload a video and specify what you'd like to add to create a seamlessly integrated result.", + category="api node/video/Pika", + inputs=[ + comfy_io.Video.Input("video", tooltip="The video to add an image to."), + comfy_io.Image.Input("image", tooltip="The image to add to the video."), + comfy_io.String.Input("prompt_text", multiline=True), + comfy_io.String.Input("negative_prompt", multiline=True), + comfy_io.Int.Input( "seed", min=0, max=0xFFFFFFFF, control_after_generate=True, ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Add any object or image into your video. Upload a video and specify what you'd like to add to create a seamlessly integrated result." - - async def api_call( - self, + @classmethod + async def execute( + cls, video: VideoInput, image: torch.Tensor, prompt_text: str, negative_prompt: str, seed: int, - unique_id: str, - **kwargs, - ) -> tuple[VideoFromFile]: + ) -> comfy_io.NodeOutput: # Convert video to BytesIO - video_bytes_io = io.BytesIO() + video_bytes_io = BytesIO() video.save_to(video_bytes_io, format=VideoContainer.MP4, codec=VideoCodec.H264) video_bytes_io.seek(0) @@ -492,7 +497,10 @@ class PikAdditionsNode(PikaNodeBase): negativePrompt=negative_prompt, seed=seed, ) - + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_PIKADDITIONS, @@ -503,74 +511,51 @@ class PikAdditionsNode(PikaNodeBase): request=pika_request_data, files=pika_files, content_type="multipart/form-data", - auth_kwargs=kwargs, + auth_kwargs=auth, ) - return await self.execute_task(initial_operation, auth_kwargs=kwargs, node_id=unique_id) + return await execute_task(initial_operation, auth_kwargs=auth, node_id=cls.hidden.unique_id) -class PikaSwapsNode(PikaNodeBase): +class PikaSwapsNode(comfy_io.ComfyNode): """Pika Pikaswaps Node.""" @classmethod - def INPUT_TYPES(cls): - return { - "required": { - "video": (IO.VIDEO, {"tooltip": "The video to swap an object in."}), - "image": ( - IO.IMAGE, - { - "tooltip": "The image used to replace the masked object in the video." - }, - ), - "mask": ( - IO.MASK, - {"tooltip": "Use the mask to define areas in the video to replace"}, - ), - "prompt_text": model_field_to_node_input( - IO.STRING, - PikaBodyGeneratePikaswapsGeneratePikaswapsPost, - "promptText", - multiline=True, - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - PikaBodyGeneratePikaswapsGeneratePikaswapsPost, - "negativePrompt", - multiline=True, - ), - "seed": model_field_to_node_input( - IO.INT, - PikaBodyGeneratePikaswapsGeneratePikaswapsPost, - "seed", - min=0, - max=0xFFFFFFFF, - control_after_generate=True, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="Pikaswaps", + display_name="Pika Swaps (Video Object Replacement)", + description="Swap out any object or region of your video with a new image or object. Define areas to replace either with a mask or coordinates.", + category="api node/video/Pika", + inputs=[ + comfy_io.Video.Input("video", tooltip="The video to swap an object in."), + comfy_io.Image.Input("image", tooltip="The image used to replace the masked object in the video."), + comfy_io.Mask.Input("mask", tooltip="Use the mask to define areas in the video to replace"), + comfy_io.String.Input("prompt_text", multiline=True), + comfy_io.String.Input("negative_prompt", multiline=True), + comfy_io.Int.Input("seed", min=0, max=0xFFFFFFFF, control_after_generate=True), + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Swap out any object or region of your video with a new image or object. Define areas to replace either with a mask or coordinates." - RETURN_TYPES = ("VIDEO",) - - async def api_call( - self, + @classmethod + async def execute( + cls, video: VideoInput, image: torch.Tensor, mask: torch.Tensor, prompt_text: str, negative_prompt: str, seed: int, - unique_id: str, - **kwargs, - ) -> tuple[VideoFromFile]: + ) -> comfy_io.NodeOutput: # Convert video to BytesIO - video_bytes_io = io.BytesIO() + video_bytes_io = BytesIO() video.save_to(video_bytes_io, format=VideoContainer.MP4, codec=VideoCodec.H264) video_bytes_io.seek(0) @@ -579,7 +564,7 @@ class PikaSwapsNode(PikaNodeBase): mask = mask.repeat(1, 3, 1, 1) # Convert 3-channel binary mask to BytesIO - mask_bytes_io = io.BytesIO() + mask_bytes_io = BytesIO() mask_bytes_io.write(mask.numpy().astype(np.uint8)) mask_bytes_io.seek(0) @@ -599,7 +584,10 @@ class PikaSwapsNode(PikaNodeBase): negativePrompt=negative_prompt, seed=seed, ) - + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_PIKADDITIONS, @@ -610,71 +598,52 @@ class PikaSwapsNode(PikaNodeBase): request=pika_request_data, files=pika_files, content_type="multipart/form-data", - auth_kwargs=kwargs, + auth_kwargs=auth, ) - - return await self.execute_task(initial_operation, auth_kwargs=kwargs, node_id=unique_id) + return await execute_task(initial_operation, auth_kwargs=auth, node_id=cls.hidden.unique_id) -class PikaffectsNode(PikaNodeBase): +class PikaffectsNode(comfy_io.ComfyNode): """Pika Pikaffects Node.""" @classmethod - def INPUT_TYPES(cls): - return { - "required": { - "image": ( - IO.IMAGE, - {"tooltip": "The reference image to apply the Pikaffect to."}, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="Pikaffects", + display_name="Pikaffects (Video Effects)", + description="Generate a video with a specific Pikaffect. Supported Pikaffects: Cake-ify, Crumble, Crush, Decapitate, Deflate, Dissolve, Explode, Eye-pop, Inflate, Levitate, Melt, Peel, Poke, Squish, Ta-da, Tear", + category="api node/video/Pika", + inputs=[ + comfy_io.Image.Input("image", tooltip="The reference image to apply the Pikaffect to."), + comfy_io.Combo.Input( + "pikaffect", options=[pikaffect.value for pikaffect in Pikaffect], default="Cake-ify" ), - "pikaffect": model_field_to_node_input( - IO.COMBO, - PikaBodyGeneratePikaffectsGeneratePikaffectsPost, - "pikaffect", - enum_type=Pikaffect, - default="Cake-ify", - ), - "prompt_text": model_field_to_node_input( - IO.STRING, - PikaBodyGeneratePikaffectsGeneratePikaffectsPost, - "promptText", - multiline=True, - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - PikaBodyGeneratePikaffectsGeneratePikaffectsPost, - "negativePrompt", - multiline=True, - ), - "seed": model_field_to_node_input( - IO.INT, - PikaBodyGeneratePikaffectsGeneratePikaffectsPost, - "seed", - min=0, - max=0xFFFFFFFF, - control_after_generate=True, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + comfy_io.String.Input("prompt_text", multiline=True), + comfy_io.String.Input("negative_prompt", multiline=True), + comfy_io.Int.Input("seed", min=0, max=0xFFFFFFFF, control_after_generate=True), + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Generate a video with a specific Pikaffect. Supported Pikaffects: Cake-ify, Crumble, Crush, Decapitate, Deflate, Dissolve, Explode, Eye-pop, Inflate, Levitate, Melt, Peel, Poke, Squish, Ta-da, Tear" - - async def api_call( - self, + @classmethod + async def execute( + cls, image: torch.Tensor, pikaffect: str, prompt_text: str, negative_prompt: str, seed: int, - unique_id: str, - **kwargs, - ) -> tuple[VideoFromFile]: - + ) -> comfy_io.NodeOutput: + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_PIKAFFECTS, @@ -690,36 +659,38 @@ class PikaffectsNode(PikaNodeBase): ), files={"image": ("image.png", tensor_to_bytesio(image), "image/png")}, content_type="multipart/form-data", - auth_kwargs=kwargs, + auth_kwargs=auth, ) - - return await self.execute_task(initial_operation, auth_kwargs=kwargs, node_id=unique_id) + return await execute_task(initial_operation, auth_kwargs=auth, node_id=cls.hidden.unique_id) -class PikaStartEndFrameNode2_2(PikaNodeBase): +class PikaStartEndFrameNode2_2(comfy_io.ComfyNode): """PikaFrames v2.2 Node.""" @classmethod - def INPUT_TYPES(cls): - return { - "required": { - "image_start": (IO.IMAGE, {"tooltip": "The first image to combine."}), - "image_end": (IO.IMAGE, {"tooltip": "The last image to combine."}), - **cls.get_base_inputs_types( - PikaBodyGenerate22KeyframeGenerate22PikaframesPost - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="PikaStartEndFrameNode2_2", + display_name="Pika Start and End Frame to Video", + description="Generate a video by combining your first and last frame. Upload two images to define the start and end points, and let the AI create a smooth transition between them.", + category="api node/video/Pika", + inputs=[ + comfy_io.Image.Input("image_start", tooltip="The first image to combine."), + comfy_io.Image.Input("image_end", tooltip="The last image to combine."), + *get_base_inputs_types(), + ], + outputs=[comfy_io.Video.Output()], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Generate a video by combining your first and last frame. Upload two images to define the start and end points, and let the AI create a smooth transition between them." - - async def api_call( - self, + @classmethod + async def execute( + cls, image_start: torch.Tensor, image_end: torch.Tensor, prompt_text: str, @@ -727,15 +698,15 @@ class PikaStartEndFrameNode2_2(PikaNodeBase): seed: int, resolution: str, duration: int, - unique_id: str, - **kwargs, - ) -> tuple[VideoFromFile]: - + ) -> comfy_io.NodeOutput: pika_files = [ ("keyFrames", ("image_start.png", tensor_to_bytesio(image_start), "image/png")), ("keyFrames", ("image_end.png", tensor_to_bytesio(image_end), "image/png")), ] - + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_PIKAFRAMES, @@ -752,28 +723,24 @@ class PikaStartEndFrameNode2_2(PikaNodeBase): ), files=pika_files, content_type="multipart/form-data", - auth_kwargs=kwargs, + auth_kwargs=auth, ) - - return await self.execute_task(initial_operation, auth_kwargs=kwargs, node_id=unique_id) + return await execute_task(initial_operation, auth_kwargs=auth, node_id=cls.hidden.unique_id) -NODE_CLASS_MAPPINGS = { - "PikaImageToVideoNode2_2": PikaImageToVideoV2_2, - "PikaTextToVideoNode2_2": PikaTextToVideoNodeV2_2, - "PikaScenesV2_2": PikaScenesV2_2, - "Pikadditions": PikAdditionsNode, - "Pikaswaps": PikaSwapsNode, - "Pikaffects": PikaffectsNode, - "PikaStartEndFrameNode2_2": PikaStartEndFrameNode2_2, -} +class PikaApiNodesExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: + return [ + PikaImageToVideoV2_2, + PikaTextToVideoNodeV2_2, + PikaScenesV2_2, + PikAdditionsNode, + PikaSwapsNode, + PikaffectsNode, + PikaStartEndFrameNode2_2, + ] -NODE_DISPLAY_NAME_MAPPINGS = { - "PikaImageToVideoNode2_2": "Pika Image to Video", - "PikaTextToVideoNode2_2": "Pika Text to Video", - "PikaScenesV2_2": "Pika Scenes (Video Image Composition)", - "Pikadditions": "Pikadditions (Video Object Insertion)", - "Pikaswaps": "Pika Swaps (Video Object Replacement)", - "Pikaffects": "Pikaffects (Video Effects)", - "PikaStartEndFrameNode2_2": "Pika Start and End Frame to Video", -} + +async def comfy_entrypoint() -> PikaApiNodesExtension: + return PikaApiNodesExtension() From 8c1991042795d06c7ccfd5d1931eb994044c75ef Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Tue, 7 Oct 2025 02:26:52 +0300 Subject: [PATCH 106/114] convert nodes_kling.py to V3 schema (#10236) --- comfy_api_nodes/nodes_kling.py | 2146 +++++++++++++++----------------- 1 file changed, 1032 insertions(+), 1114 deletions(-) diff --git a/comfy_api_nodes/nodes_kling.py b/comfy_api_nodes/nodes_kling.py index 44fccc0c7..457b43451 100644 --- a/comfy_api_nodes/nodes_kling.py +++ b/comfy_api_nodes/nodes_kling.py @@ -10,6 +10,8 @@ from collections.abc import Callable import math import logging +from typing_extensions import override + import torch from comfy_api_nodes.apis import ( @@ -63,8 +65,8 @@ from comfy_api_nodes.apinode_utils import ( upload_video_to_comfyapi, upload_audio_to_comfyapi, download_url_to_image_tensor, + validate_string, ) -from comfy_api_nodes.mapper_utils import model_field_to_node_input from comfy_api_nodes.util.validation_utils import ( validate_image_dimensions, validate_image_aspect_ratio, @@ -73,8 +75,7 @@ from comfy_api_nodes.util.validation_utils import ( ) from comfy_api.input.basic_types import AudioInput from comfy_api.input.video_types import VideoInput -from comfy_api.input_impl import VideoFromFile -from comfy.comfy_types.node_typing import IO, InputTypeOptions, ComfyNodeABC +from comfy_api.latest import ComfyExtension, io as comfy_io KLING_API_VERSION = "v1" PATH_TEXT_TO_VIDEO = f"/proxy/kling/{KLING_API_VERSION}/videos/text2video" @@ -103,10 +104,113 @@ AVERAGE_DURATION_VIDEO_EXTEND = 320 R = TypeVar("R") -class KlingApiError(Exception): - """Base exception for Kling API errors.""" +MODE_TEXT2VIDEO = { + "standard mode / 5s duration / kling-v1": ("std", "5", "kling-v1"), + "standard mode / 10s duration / kling-v1": ("std", "10", "kling-v1"), + "pro mode / 5s duration / kling-v1": ("pro", "5", "kling-v1"), + "pro mode / 10s duration / kling-v1": ("pro", "10", "kling-v1"), + "standard mode / 5s duration / kling-v1-6": ("std", "5", "kling-v1-6"), + "standard mode / 10s duration / kling-v1-6": ("std", "10", "kling-v1-6"), + "pro mode / 5s duration / kling-v2-master": ("pro", "5", "kling-v2-master"), + "pro mode / 10s duration / kling-v2-master": ("pro", "10", "kling-v2-master"), + "standard mode / 5s duration / kling-v2-master": ("std", "5", "kling-v2-master"), + "standard mode / 10s duration / kling-v2-master": ("std", "10", "kling-v2-master"), + "pro mode / 5s duration / kling-v2-1-master": ("pro", "5", "kling-v2-1-master"), + "pro mode / 10s duration / kling-v2-1-master": ("pro", "10", "kling-v2-1-master"), + "pro mode / 5s duration / kling-v2-5-turbo": ("pro", "5", "kling-v2-5-turbo"), + "pro mode / 10s duration / kling-v2-5-turbo": ("pro", "10", "kling-v2-5-turbo"), +} +""" +Mapping of mode strings to their corresponding (mode, duration, model_name) tuples. +Only includes config combos that support the `image_tail` request field. - pass +See: [Kling API Docs Capability Map](https://app.klingai.com/global/dev/document-api/apiReference/model/skillsMap) +""" + + +MODE_START_END_FRAME = { + "standard mode / 5s duration / kling-v1": ("std", "5", "kling-v1"), + "pro mode / 5s duration / kling-v1": ("pro", "5", "kling-v1"), + "pro mode / 5s duration / kling-v1-5": ("pro", "5", "kling-v1-5"), + "pro mode / 10s duration / kling-v1-5": ("pro", "10", "kling-v1-5"), + "pro mode / 5s duration / kling-v1-6": ("pro", "5", "kling-v1-6"), + "pro mode / 10s duration / kling-v1-6": ("pro", "10", "kling-v1-6"), + "pro mode / 5s duration / kling-v2-1": ("pro", "5", "kling-v2-1"), + "pro mode / 10s duration / kling-v2-1": ("pro", "10", "kling-v2-1"), +} +""" +Returns a mapping of mode strings to their corresponding (mode, duration, model_name) tuples. +Only includes config combos that support the `image_tail` request field. + +See: [Kling API Docs Capability Map](https://app.klingai.com/global/dev/document-api/apiReference/model/skillsMap) +""" + + +VOICES_CONFIG = { + # English voices + "Melody": ("girlfriend_4_speech02", "en"), + "Sunny": ("genshin_vindi2", "en"), + "Sage": ("zhinen_xuesheng", "en"), + "Ace": ("AOT", "en"), + "Blossom": ("ai_shatang", "en"), + "Peppy": ("genshin_klee2", "en"), + "Dove": ("genshin_kirara", "en"), + "Shine": ("ai_kaiya", "en"), + "Anchor": ("oversea_male1", "en"), + "Lyric": ("ai_chenjiahao_712", "en"), + "Tender": ("chat1_female_new-3", "en"), + "Siren": ("chat_0407_5-1", "en"), + "Zippy": ("cartoon-boy-07", "en"), + "Bud": ("uk_boy1", "en"), + "Sprite": ("cartoon-girl-01", "en"), + "Candy": ("PeppaPig_platform", "en"), + "Beacon": ("ai_huangzhong_712", "en"), + "Rock": ("ai_huangyaoshi_712", "en"), + "Titan": ("ai_laoguowang_712", "en"), + "Grace": ("chengshu_jiejie", "en"), + "Helen": ("you_pingjing", "en"), + "Lore": ("calm_story1", "en"), + "Crag": ("uk_man2", "en"), + "Prattle": ("laopopo_speech02", "en"), + "Hearth": ("heainainai_speech02", "en"), + "The Reader": ("reader_en_m-v1", "en"), + "Commercial Lady": ("commercial_lady_en_f-v1", "en"), + # Chinese voices + "阳光少年": ("genshin_vindi2", "zh"), + "懂事小弟": ("zhinen_xuesheng", "zh"), + "运动少年": ("tiyuxi_xuedi", "zh"), + "青春少女": ("ai_shatang", "zh"), + "温柔小妹": ("genshin_klee2", "zh"), + "元气少女": ("genshin_kirara", "zh"), + "阳光男生": ("ai_kaiya", "zh"), + "幽默小哥": ("tiexin_nanyou", "zh"), + "文艺小哥": ("ai_chenjiahao_712", "zh"), + "甜美邻家": ("girlfriend_1_speech02", "zh"), + "温柔姐姐": ("chat1_female_new-3", "zh"), + "职场女青": ("girlfriend_2_speech02", "zh"), + "活泼男童": ("cartoon-boy-07", "zh"), + "俏皮女童": ("cartoon-girl-01", "zh"), + "稳重老爸": ("ai_huangyaoshi_712", "zh"), + "温柔妈妈": ("you_pingjing", "zh"), + "严肃上司": ("ai_laoguowang_712", "zh"), + "优雅贵妇": ("chengshu_jiejie", "zh"), + "慈祥爷爷": ("zhuxi_speech02", "zh"), + "唠叨爷爷": ("uk_oldman3", "zh"), + "唠叨奶奶": ("laopopo_speech02", "zh"), + "和蔼奶奶": ("heainainai_speech02", "zh"), + "东北老铁": ("dongbeilaotie_speech02", "zh"), + "重庆小伙": ("chongqingxiaohuo_speech02", "zh"), + "四川妹子": ("chuanmeizi_speech02", "zh"), + "潮汕大叔": ("chaoshandashu_speech02", "zh"), + "台湾男生": ("ai_taiwan_man2_speech02", "zh"), + "西安掌柜": ("xianzhanggui_speech02", "zh"), + "天津姐姐": ("tianjinjiejie_speech02", "zh"), + "新闻播报男": ("diyinnansang_DB_CN_M_04-v2", "zh"), + "译制片男": ("yizhipiannan-v1", "zh"), + "撒娇女友": ("tianmeixuemei-v1", "zh"), + "刀片烟嗓": ("daopianyansang-v1", "zh"), + "乖巧正太": ("mengwa-v1", "zh"), +} async def poll_until_finished( @@ -142,11 +246,6 @@ def is_valid_camera_control_configs(configs: list[float]) -> bool: return any(not math.isclose(value, 0.0) for value in configs) -def is_valid_prompt(prompt: str) -> bool: - """Verifies that the prompt is not empty.""" - return bool(prompt) - - def is_valid_task_creation_response(response: KlingText2VideoResponse) -> bool: """Verifies that the initial response contains a task ID.""" return bool(response.data.task_id) @@ -190,7 +289,7 @@ def validate_task_creation_response(response) -> None: if not is_valid_task_creation_response(response): error_msg = f"Kling initial request failed. Code: {response.code}, Message: {response.message}, Data: {response.data}" logging.error(error_msg) - raise KlingApiError(error_msg) + raise Exception(error_msg) def validate_video_result_response(response) -> None: @@ -198,7 +297,7 @@ def validate_video_result_response(response) -> None: if not is_valid_video_response(response): error_msg = f"Kling task {response.data.task_id} succeeded but no video data found in response." logging.error(f"Error: {error_msg}.\nResponse: {response}") - raise KlingApiError(error_msg) + raise Exception(error_msg) def validate_image_result_response(response) -> None: @@ -206,7 +305,7 @@ def validate_image_result_response(response) -> None: if not is_valid_image_response(response): error_msg = f"Kling task {response.data.task_id} succeeded but no image data found in response." logging.error(f"Error: {error_msg}.\nResponse: {response}") - raise KlingApiError(error_msg) + raise Exception(error_msg) def validate_input_image(image: torch.Tensor) -> None: @@ -221,21 +320,6 @@ def validate_input_image(image: torch.Tensor) -> None: validate_image_aspect_ratio(image, min_aspect_ratio=1 / 2.5, max_aspect_ratio=2.5) -def get_camera_control_input_config( - tooltip: str, default: float = 0.0 -) -> tuple[IO, InputTypeOptions]: - """Returns common InputTypeOptions for Kling camera control configurations.""" - input_config = { - "default": default, - "min": -10.0, - "max": 10.0, - "step": 0.25, - "display": "slider", - "tooltip": tooltip, - } - return IO.FLOAT, input_config - - def get_video_from_response(response) -> KlingVideoResult: """Returns the first video object from the Kling video generation task result. Will raise an error if the response is not valid. @@ -278,17 +362,6 @@ def get_images_urls_from_response(response) -> Optional[str]: return None -async def video_result_to_node_output( - video: KlingVideoResult, -) -> tuple[VideoFromFile, str, str]: - """Converts a KlingVideoResult to a tuple of (VideoFromFile, str, str) to be used as a ComfyUI node output.""" - return ( - await download_url_to_video_output(str(video.url)), - str(video.id), - str(video.duration), - ) - - async def image_result_to_node_output( images: list[KlingImageResult], ) -> torch.Tensor: @@ -302,57 +375,339 @@ async def image_result_to_node_output( return torch.cat([await download_url_to_image_tensor(str(image.url)) for image in images]) -class KlingNodeBase(ComfyNodeABC): - """Base class for Kling nodes.""" +async def execute_text2video( + auth_kwargs: dict[str, str], + node_id: str, + prompt: str, + negative_prompt: str, + cfg_scale: float, + model_name: str, + model_mode: str, + duration: str, + aspect_ratio: str, + camera_control: Optional[KlingCameraControl] = None, +) -> comfy_io.NodeOutput: + validate_prompts(prompt, negative_prompt, MAX_PROMPT_LENGTH_T2V) + initial_operation = SynchronousOperation( + endpoint=ApiEndpoint( + path=PATH_TEXT_TO_VIDEO, + method=HttpMethod.POST, + request_model=KlingText2VideoRequest, + response_model=KlingText2VideoResponse, + ), + request=KlingText2VideoRequest( + prompt=prompt if prompt else None, + negative_prompt=negative_prompt if negative_prompt else None, + duration=KlingVideoGenDuration(duration), + mode=KlingVideoGenMode(model_mode), + model_name=KlingVideoGenModelName(model_name), + cfg_scale=cfg_scale, + aspect_ratio=KlingVideoGenAspectRatio(aspect_ratio), + camera_control=camera_control, + ), + auth_kwargs=auth_kwargs, + ) - FUNCTION = "api_call" - CATEGORY = "api node/video/Kling" - API_NODE = True + task_creation_response = await initial_operation.execute() + validate_task_creation_response(task_creation_response) + + task_id = task_creation_response.data.task_id + final_response = await poll_until_finished( + auth_kwargs, + ApiEndpoint( + path=f"{PATH_TEXT_TO_VIDEO}/{task_id}", + method=HttpMethod.GET, + request_model=EmptyRequest, + response_model=KlingText2VideoResponse, + ), + result_url_extractor=get_video_url_from_response, + estimated_duration=AVERAGE_DURATION_T2V, + node_id=node_id, + ) + validate_video_result_response(final_response) + + video = get_video_from_response(final_response) + return comfy_io.NodeOutput(await download_url_to_video_output(str(video.url)), str(video.id), str(video.duration)) -class KlingCameraControls(KlingNodeBase): +async def execute_image2video( + auth_kwargs: dict[str, str], + node_id: str, + start_frame: torch.Tensor, + prompt: str, + negative_prompt: str, + model_name: str, + cfg_scale: float, + model_mode: str, + aspect_ratio: str, + duration: str, + camera_control: Optional[KlingCameraControl] = None, + end_frame: Optional[torch.Tensor] = None, +) -> comfy_io.NodeOutput: + validate_prompts(prompt, negative_prompt, MAX_PROMPT_LENGTH_I2V) + validate_input_image(start_frame) + + if camera_control is not None: + # Camera control type for image 2 video is always `simple` + camera_control.type = KlingCameraControlType.simple + + if model_mode == "std" and model_name == KlingVideoGenModelName.kling_v2_5_turbo.value: + model_mode = "pro" # October 5: currently "std" mode is not supported for this model + + initial_operation = SynchronousOperation( + endpoint=ApiEndpoint( + path=PATH_IMAGE_TO_VIDEO, + method=HttpMethod.POST, + request_model=KlingImage2VideoRequest, + response_model=KlingImage2VideoResponse, + ), + request=KlingImage2VideoRequest( + model_name=KlingVideoGenModelName(model_name), + image=tensor_to_base64_string(start_frame), + image_tail=( + tensor_to_base64_string(end_frame) + if end_frame is not None + else None + ), + prompt=prompt, + negative_prompt=negative_prompt if negative_prompt else None, + cfg_scale=cfg_scale, + mode=KlingVideoGenMode(model_mode), + duration=KlingVideoGenDuration(duration), + camera_control=camera_control, + ), + auth_kwargs=auth_kwargs, + ) + + task_creation_response = await initial_operation.execute() + validate_task_creation_response(task_creation_response) + task_id = task_creation_response.data.task_id + + final_response = await poll_until_finished( + auth_kwargs, + ApiEndpoint( + path=f"{PATH_IMAGE_TO_VIDEO}/{task_id}", + method=HttpMethod.GET, + request_model=KlingImage2VideoRequest, + response_model=KlingImage2VideoResponse, + ), + result_url_extractor=get_video_url_from_response, + estimated_duration=AVERAGE_DURATION_I2V, + node_id=node_id, + ) + validate_video_result_response(final_response) + + video = get_video_from_response(final_response) + return comfy_io.NodeOutput(await download_url_to_video_output(str(video.url)), str(video.id), str(video.duration)) + + +async def execute_video_effect( + auth_kwargs: dict[str, str], + node_id: str, + dual_character: bool, + effect_scene: KlingDualCharacterEffectsScene | KlingSingleImageEffectsScene, + model_name: str, + duration: KlingVideoGenDuration, + image_1: torch.Tensor, + image_2: Optional[torch.Tensor] = None, + model_mode: Optional[KlingVideoGenMode] = None, +) -> comfy_io.NodeOutput: + if dual_character: + request_input_field = KlingDualCharacterEffectInput( + model_name=model_name, + mode=model_mode, + images=[ + tensor_to_base64_string(image_1), + tensor_to_base64_string(image_2), + ], + duration=duration, + ) + else: + request_input_field = KlingSingleImageEffectInput( + model_name=model_name, + image=tensor_to_base64_string(image_1), + duration=duration, + ) + + initial_operation = SynchronousOperation( + endpoint=ApiEndpoint( + path=PATH_VIDEO_EFFECTS, + method=HttpMethod.POST, + request_model=KlingVideoEffectsRequest, + response_model=KlingVideoEffectsResponse, + ), + request=KlingVideoEffectsRequest( + effect_scene=effect_scene, + input=request_input_field, + ), + auth_kwargs=auth_kwargs, + ) + + task_creation_response = await initial_operation.execute() + validate_task_creation_response(task_creation_response) + task_id = task_creation_response.data.task_id + + final_response = await poll_until_finished( + auth_kwargs, + ApiEndpoint( + path=f"{PATH_VIDEO_EFFECTS}/{task_id}", + method=HttpMethod.GET, + request_model=EmptyRequest, + response_model=KlingVideoEffectsResponse, + ), + result_url_extractor=get_video_url_from_response, + estimated_duration=AVERAGE_DURATION_VIDEO_EFFECTS, + node_id=node_id, + ) + validate_video_result_response(final_response) + + video = get_video_from_response(final_response) + return comfy_io.NodeOutput(await download_url_to_video_output(str(video.url)), str(video.id), str(video.duration)) + + +async def execute_lipsync( + auth_kwargs: dict[str, str], + node_id: str, + video: VideoInput, + audio: Optional[AudioInput] = None, + voice_language: Optional[str] = None, + model_mode: Optional[str] = None, + text: Optional[str] = None, + voice_speed: Optional[float] = None, + voice_id: Optional[str] = None, +) -> comfy_io.NodeOutput: + if text: + validate_string(text, field_name="Text", max_length=MAX_PROMPT_LENGTH_LIP_SYNC) + validate_video_dimensions(video, 720, 1920) + validate_video_duration(video, 2, 10) + + # Upload video to Comfy API and get download URL + video_url = await upload_video_to_comfyapi(video, auth_kwargs=auth_kwargs) + logging.info("Uploaded video to Comfy API. URL: %s", video_url) + + # Upload the audio file to Comfy API and get download URL + if audio: + audio_url = await upload_audio_to_comfyapi(audio, auth_kwargs=auth_kwargs) + logging.info("Uploaded audio to Comfy API. URL: %s", audio_url) + else: + audio_url = None + + initial_operation = SynchronousOperation( + endpoint=ApiEndpoint( + path=PATH_LIP_SYNC, + method=HttpMethod.POST, + request_model=KlingLipSyncRequest, + response_model=KlingLipSyncResponse, + ), + request=KlingLipSyncRequest( + input=KlingLipSyncInputObject( + video_url=video_url, + mode=model_mode, + text=text, + voice_language=voice_language, + voice_speed=voice_speed, + audio_type="url", + audio_url=audio_url, + voice_id=voice_id, + ), + ), + auth_kwargs=auth_kwargs, + ) + + task_creation_response = await initial_operation.execute() + validate_task_creation_response(task_creation_response) + task_id = task_creation_response.data.task_id + + final_response = await poll_until_finished( + auth_kwargs, + ApiEndpoint( + path=f"{PATH_LIP_SYNC}/{task_id}", + method=HttpMethod.GET, + request_model=EmptyRequest, + response_model=KlingLipSyncResponse, + ), + result_url_extractor=get_video_url_from_response, + estimated_duration=AVERAGE_DURATION_LIP_SYNC, + node_id=node_id, + ) + validate_video_result_response(final_response) + + video = get_video_from_response(final_response) + return comfy_io.NodeOutput(await download_url_to_video_output(str(video.url)), str(video.id), str(video.duration)) + + +class KlingCameraControls(comfy_io.ComfyNode): """Kling Camera Controls Node""" @classmethod - def INPUT_TYPES(cls): - return { - "required": { - "camera_control_type": model_field_to_node_input( - IO.COMBO, - KlingCameraControl, - "type", - enum_type=KlingCameraControlType, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingCameraControls", + display_name="Kling Camera Controls", + category="api node/video/Kling", + description="Allows specifying configuration options for Kling Camera Controls and motion control effects.", + inputs=[ + comfy_io.Combo.Input("camera_control_type", options=[i.value for i in KlingCameraControlType]), + comfy_io.Float.Input( + "horizontal_movement", + default=0.0, + min=-10.0, + max=10.0, + step=0.25, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Controls camera's movement along horizontal axis (x-axis). Negative indicates left, positive indicates right", ), - "horizontal_movement": get_camera_control_input_config( - "Controls camera's movement along horizontal axis (x-axis). Negative indicates left, positive indicates right" + comfy_io.Float.Input( + "vertical_movement", + default=0.0, + min=-10.0, + max=10.0, + step=0.25, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Controls camera's movement along vertical axis (y-axis). Negative indicates downward, positive indicates upward.", ), - "vertical_movement": get_camera_control_input_config( - "Controls camera's movement along vertical axis (y-axis). Negative indicates downward, positive indicates upward." - ), - "pan": get_camera_control_input_config( - "Controls camera's rotation in vertical plane (x-axis). Negative indicates downward rotation, positive indicates upward rotation.", + comfy_io.Float.Input( + "pan", default=0.5, + min=-10.0, + max=10.0, + step=0.25, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Controls camera's rotation in vertical plane (x-axis). Negative indicates downward rotation, positive indicates upward rotation.", ), - "tilt": get_camera_control_input_config( - "Controls camera's rotation in horizontal plane (y-axis). Negative indicates left rotation, positive indicates right rotation.", + comfy_io.Float.Input( + "tilt", + default=0.0, + min=-10.0, + max=10.0, + step=0.25, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Controls camera's rotation in horizontal plane (y-axis). Negative indicates left rotation, positive indicates right rotation.", ), - "roll": get_camera_control_input_config( - "Controls camera's rolling amount (z-axis). Negative indicates counterclockwise, positive indicates clockwise.", + comfy_io.Float.Input( + "roll", + default=0.0, + min=-10.0, + max=10.0, + step=0.25, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Controls camera's rolling amount (z-axis). Negative indicates counterclockwise, positive indicates clockwise.", ), - "zoom": get_camera_control_input_config( - "Controls change in camera's focal length. Negative indicates narrower field of view, positive indicates wider field of view.", + comfy_io.Float.Input( + "zoom", + default=0.0, + min=-10.0, + max=10.0, + step=0.25, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Controls change in camera's focal length. Negative indicates narrower field of view, positive indicates wider field of view.", ), - } - } - - DESCRIPTION = "Allows specifying configuration options for Kling Camera Controls and motion control effects." - RETURN_TYPES = ("CAMERA_CONTROL",) - RETURN_NAMES = ("camera_control",) - FUNCTION = "main" - API_NODE = False # This is just a helper node, it doesn't make an API call + ], + outputs=[comfy_io.Custom("CAMERA_CONTROL").Output(display_name="camera_control")], + ) @classmethod - def VALIDATE_INPUTS( + def validate_inputs( cls, horizontal_movement: float, vertical_movement: float, @@ -374,8 +729,9 @@ class KlingCameraControls(KlingNodeBase): return "Invalid camera control configs: at least one of the values must be non-zero" return True - def main( - self, + @classmethod + def execute( + cls, camera_control_type: str, horizontal_movement: float, vertical_movement: float, @@ -383,8 +739,8 @@ class KlingCameraControls(KlingNodeBase): tilt: float, roll: float, zoom: float, - ) -> tuple[KlingCameraControl]: - return ( + ) -> comfy_io.NodeOutput: + return comfy_io.NodeOutput( KlingCameraControl( type=KlingCameraControlType(camera_control_type), config=KlingCameraConfig( @@ -395,303 +751,186 @@ class KlingCameraControls(KlingNodeBase): tilt=tilt, zoom=zoom, ), - ), + ) ) -class KlingTextToVideoNode(KlingNodeBase): +class KlingTextToVideoNode(comfy_io.ComfyNode): """Kling Text to Video Node""" - @staticmethod - def get_mode_string_mapping() -> dict[str, tuple[str, str, str]]: - """ - Returns a mapping of mode strings to their corresponding (mode, duration, model_name) tuples. - Only includes config combos that support the `image_tail` request field. - - See: [Kling API Docs Capability Map](https://app.klingai.com/global/dev/document-api/apiReference/model/skillsMap) - """ - return { - "standard mode / 5s duration / kling-v1": ("std", "5", "kling-v1"), - "standard mode / 10s duration / kling-v1": ("std", "10", "kling-v1"), - "pro mode / 5s duration / kling-v1": ("pro", "5", "kling-v1"), - "pro mode / 10s duration / kling-v1": ("pro", "10", "kling-v1"), - "standard mode / 5s duration / kling-v1-6": ("std", "5", "kling-v1-6"), - "standard mode / 10s duration / kling-v1-6": ("std", "10", "kling-v1-6"), - "pro mode / 5s duration / kling-v2-master": ("pro", "5", "kling-v2-master"), - "pro mode / 10s duration / kling-v2-master": ("pro", "10", "kling-v2-master"), - "standard mode / 5s duration / kling-v2-master": ("std", "5", "kling-v2-master"), - "standard mode / 10s duration / kling-v2-master": ("std", "10", "kling-v2-master"), - "pro mode / 5s duration / kling-v2-1-master": ("pro", "5", "kling-v2-1-master"), - "pro mode / 10s duration / kling-v2-1-master": ("pro", "10", "kling-v2-1-master"), - "pro mode / 5s duration / kling-v2-5-turbo": ("pro", "5", "kling-v2-5-turbo"), - "pro mode / 10s duration / kling-v2-5-turbo": ("pro", "10", "kling-v2-5-turbo"), - } - @classmethod - def INPUT_TYPES(s): - modes = list(KlingTextToVideoNode.get_mode_string_mapping().keys()) - return { - "required": { - "prompt": model_field_to_node_input( - IO.STRING, KlingText2VideoRequest, "prompt", multiline=True - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, KlingText2VideoRequest, "negative_prompt", multiline=True - ), - "cfg_scale": model_field_to_node_input( - IO.FLOAT, - KlingText2VideoRequest, - "cfg_scale", - default=1.0, - min=0.0, - max=1.0, - ), - "aspect_ratio": model_field_to_node_input( - IO.COMBO, - KlingText2VideoRequest, + def define_schema(cls) -> comfy_io.Schema: + modes = list(MODE_TEXT2VIDEO.keys()) + return comfy_io.Schema( + node_id="KlingTextToVideoNode", + display_name="Kling Text to Video", + category="api node/video/Kling", + description="Kling Text to Video Node", + inputs=[ + comfy_io.String.Input("prompt", multiline=True, tooltip="Positive text prompt"), + comfy_io.String.Input("negative_prompt", multiline=True, tooltip="Negative text prompt"), + comfy_io.Float.Input("cfg_scale", default=1.0, min=0.0, max=1.0), + comfy_io.Combo.Input( "aspect_ratio", - enum_type=KlingVideoGenAspectRatio, + options=[i.value for i in KlingVideoGenAspectRatio], + default="16:9", ), - "mode": ( - modes, - { - "default": modes[4], - "tooltip": "The configuration to use for the video generation following the format: mode / duration / model_name.", - }, + comfy_io.Combo.Input( + "mode", + options=modes, + default=modes[4], + tooltip="The configuration to use for the video generation following the format: mode / duration / model_name.", ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - RETURN_TYPES = ("VIDEO", "STRING", "STRING") - RETURN_NAMES = ("VIDEO", "video_id", "duration") - DESCRIPTION = "Kling Text to Video Node" - - async def get_response( - self, task_id: str, auth_kwargs: dict[str, str], node_id: Optional[str] = None - ) -> KlingText2VideoResponse: - return await poll_until_finished( - auth_kwargs, - ApiEndpoint( - path=f"{PATH_TEXT_TO_VIDEO}/{task_id}", - method=HttpMethod.GET, - request_model=EmptyRequest, - response_model=KlingText2VideoResponse, - ), - result_url_extractor=get_video_url_from_response, - estimated_duration=AVERAGE_DURATION_T2V, - node_id=node_id, + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, ) - async def api_call( - self, + @classmethod + async def execute( + cls, prompt: str, negative_prompt: str, cfg_scale: float, mode: str, aspect_ratio: str, - camera_control: Optional[KlingCameraControl] = None, - model_name: Optional[str] = None, - duration: Optional[str] = None, - unique_id: Optional[str] = None, - **kwargs, - ) -> tuple[VideoFromFile, str, str]: - validate_prompts(prompt, negative_prompt, MAX_PROMPT_LENGTH_T2V) - if model_name is None: - mode, duration, model_name = self.get_mode_string_mapping()[mode] - initial_operation = SynchronousOperation( - endpoint=ApiEndpoint( - path=PATH_TEXT_TO_VIDEO, - method=HttpMethod.POST, - request_model=KlingText2VideoRequest, - response_model=KlingText2VideoResponse, - ), - request=KlingText2VideoRequest( - prompt=prompt if prompt else None, - negative_prompt=negative_prompt if negative_prompt else None, - duration=KlingVideoGenDuration(duration), - mode=KlingVideoGenMode(mode), - model_name=KlingVideoGenModelName(model_name), - cfg_scale=cfg_scale, - aspect_ratio=KlingVideoGenAspectRatio(aspect_ratio), - camera_control=camera_control, - ), - auth_kwargs=kwargs, + ) -> comfy_io.NodeOutput: + model_mode, duration, model_name = MODE_TEXT2VIDEO[mode] + return await execute_text2video( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, + prompt=prompt, + negative_prompt=negative_prompt, + cfg_scale=cfg_scale, + model_mode=model_mode, + aspect_ratio=aspect_ratio, + model_name=model_name, + duration=duration, ) - task_creation_response = await initial_operation.execute() - validate_task_creation_response(task_creation_response) - task_id = task_creation_response.data.task_id - final_response = await self.get_response( - task_id, auth_kwargs=kwargs, node_id=unique_id - ) - validate_video_result_response(final_response) - - video = get_video_from_response(final_response) - return await video_result_to_node_output(video) - - -class KlingCameraControlT2VNode(KlingTextToVideoNode): +class KlingCameraControlT2VNode(comfy_io.ComfyNode): """ Kling Text to Video Camera Control Node. This node is a text to video node, but it supports controlling the camera. Duration, mode, and model_name request fields are hard-coded because camera control is only supported in pro mode with the kling-v1-5 model at 5s duration as of 2025-05-02. """ @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": model_field_to_node_input( - IO.STRING, KlingText2VideoRequest, "prompt", multiline=True - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - KlingText2VideoRequest, - "negative_prompt", - multiline=True, - ), - "cfg_scale": model_field_to_node_input( - IO.FLOAT, - KlingText2VideoRequest, - "cfg_scale", - default=0.75, - min=0.0, - max=1.0, - ), - "aspect_ratio": model_field_to_node_input( - IO.COMBO, - KlingText2VideoRequest, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingCameraControlT2VNode", + display_name="Kling Text to Video (Camera Control)", + category="api node/video/Kling", + description="Transform text into cinematic videos with professional camera movements that simulate real-world cinematography. Control virtual camera actions including zoom, rotation, pan, tilt, and first-person view, while maintaining focus on your original text.", + inputs=[ + comfy_io.String.Input("prompt", multiline=True, tooltip="Positive text prompt"), + comfy_io.String.Input("negative_prompt", multiline=True, tooltip="Negative text prompt"), + comfy_io.Float.Input("cfg_scale", default=0.75, min=0.0, max=1.0), + comfy_io.Combo.Input( "aspect_ratio", - enum_type=KlingVideoGenAspectRatio, + options=[i.value for i in KlingVideoGenAspectRatio], + default="16:9", ), - "camera_control": ( - "CAMERA_CONTROL", - { - "tooltip": "Can be created using the Kling Camera Controls node. Controls the camera movement and motion during the video generation.", - }, + comfy_io.Custom("CAMERA_CONTROL").Input( + "camera_control", + tooltip="Can be created using the Kling Camera Controls node. Controls the camera movement and motion during the video generation.", ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Transform text into cinematic videos with professional camera movements that simulate real-world cinematography. Control virtual camera actions including zoom, rotation, pan, tilt, and first-person view, while maintaining focus on your original text." - - async def api_call( - self, + @classmethod + async def execute( + cls, prompt: str, negative_prompt: str, cfg_scale: float, aspect_ratio: str, camera_control: Optional[KlingCameraControl] = None, - unique_id: Optional[str] = None, - **kwargs, - ): - return await super().api_call( + ) -> comfy_io.NodeOutput: + return await execute_text2video( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, model_name=KlingVideoGenModelName.kling_v1, cfg_scale=cfg_scale, - mode=KlingVideoGenMode.std, + model_mode=KlingVideoGenMode.std, aspect_ratio=KlingVideoGenAspectRatio(aspect_ratio), duration=KlingVideoGenDuration.field_5, prompt=prompt, negative_prompt=negative_prompt, camera_control=camera_control, - **kwargs, ) -class KlingImage2VideoNode(KlingNodeBase): +class KlingImage2VideoNode(comfy_io.ComfyNode): """Kling Image to Video Node""" @classmethod - def INPUT_TYPES(s): - return { - "required": { - "start_frame": model_field_to_node_input( - IO.IMAGE, - KlingImage2VideoRequest, - "image", - tooltip="The reference image used to generate the video.", - ), - "prompt": model_field_to_node_input( - IO.STRING, KlingImage2VideoRequest, "prompt", multiline=True - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - KlingImage2VideoRequest, - "negative_prompt", - multiline=True, - ), - "model_name": model_field_to_node_input( - IO.COMBO, - KlingImage2VideoRequest, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingImage2VideoNode", + display_name="Kling Image to Video", + category="api node/video/Kling", + description="Kling Image to Video Node", + inputs=[ + comfy_io.Image.Input("start_frame", tooltip="The reference image used to generate the video."), + comfy_io.String.Input("prompt", multiline=True, tooltip="Positive text prompt"), + comfy_io.String.Input("negative_prompt", multiline=True, tooltip="Negative text prompt"), + comfy_io.Combo.Input( "model_name", - enum_type=KlingVideoGenModelName, + options=[i.value for i in KlingVideoGenModelName], + default="kling-v2-master", ), - "cfg_scale": model_field_to_node_input( - IO.FLOAT, - KlingImage2VideoRequest, - "cfg_scale", - default=0.8, - min=0.0, - max=1.0, - ), - "mode": model_field_to_node_input( - IO.COMBO, - KlingImage2VideoRequest, - "mode", - enum_type=KlingVideoGenMode, - ), - "aspect_ratio": model_field_to_node_input( - IO.COMBO, - KlingImage2VideoRequest, + comfy_io.Float.Input("cfg_scale", default=0.8, min=0.0, max=1.0), + comfy_io.Combo.Input("mode", options=[i.value for i in KlingVideoGenMode], default="std"), + comfy_io.Combo.Input( "aspect_ratio", - enum_type=KlingVideoGenAspectRatio, + options=[i.value for i in KlingVideoGenAspectRatio], + default="16:9", ), - "duration": model_field_to_node_input( - IO.COMBO, - KlingImage2VideoRequest, - "duration", - enum_type=KlingVideoGenDuration, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - RETURN_TYPES = ("VIDEO", "STRING", "STRING") - RETURN_NAMES = ("VIDEO", "video_id", "duration") - DESCRIPTION = "Kling Image to Video Node" - - async def get_response( - self, task_id: str, auth_kwargs: dict[str, str], node_id: Optional[str] = None - ) -> KlingImage2VideoResponse: - return await poll_until_finished( - auth_kwargs, - ApiEndpoint( - path=f"{PATH_IMAGE_TO_VIDEO}/{task_id}", - method=HttpMethod.GET, - request_model=KlingImage2VideoRequest, - response_model=KlingImage2VideoResponse, - ), - result_url_extractor=get_video_url_from_response, - estimated_duration=AVERAGE_DURATION_I2V, - node_id=node_id, + comfy_io.Combo.Input("duration", options=[i.value for i in KlingVideoGenDuration], default="5"), + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, ) - async def api_call( - self, + @classmethod + async def execute( + cls, start_frame: torch.Tensor, prompt: str, negative_prompt: str, @@ -702,212 +941,151 @@ class KlingImage2VideoNode(KlingNodeBase): duration: str, camera_control: Optional[KlingCameraControl] = None, end_frame: Optional[torch.Tensor] = None, - unique_id: Optional[str] = None, - **kwargs, - ) -> tuple[VideoFromFile]: - validate_prompts(prompt, negative_prompt, MAX_PROMPT_LENGTH_I2V) - validate_input_image(start_frame) - - if camera_control is not None: - # Camera control type for image 2 video is always `simple` - camera_control.type = KlingCameraControlType.simple - - if mode == "std" and model_name == KlingVideoGenModelName.kling_v2_5_turbo.value: - mode = "pro" # October 5: currently "std" mode is not supported for this model - - initial_operation = SynchronousOperation( - endpoint=ApiEndpoint( - path=PATH_IMAGE_TO_VIDEO, - method=HttpMethod.POST, - request_model=KlingImage2VideoRequest, - response_model=KlingImage2VideoResponse, - ), - request=KlingImage2VideoRequest( - model_name=KlingVideoGenModelName(model_name), - image=tensor_to_base64_string(start_frame), - image_tail=( - tensor_to_base64_string(end_frame) - if end_frame is not None - else None - ), - prompt=prompt, - negative_prompt=negative_prompt if negative_prompt else None, - cfg_scale=cfg_scale, - mode=KlingVideoGenMode(mode), - duration=KlingVideoGenDuration(duration), - camera_control=camera_control, - ), - auth_kwargs=kwargs, + ) -> comfy_io.NodeOutput: + return await execute_image2video( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, + start_frame=start_frame, + prompt=prompt, + negative_prompt=negative_prompt, + cfg_scale=cfg_scale, + model_name=model_name, + aspect_ratio=aspect_ratio, + model_mode=mode, + duration=duration, + camera_control=camera_control, + end_frame=end_frame, ) - task_creation_response = await initial_operation.execute() - validate_task_creation_response(task_creation_response) - task_id = task_creation_response.data.task_id - final_response = await self.get_response( - task_id, auth_kwargs=kwargs, node_id=unique_id - ) - validate_video_result_response(final_response) - - video = get_video_from_response(final_response) - return await video_result_to_node_output(video) - - -class KlingCameraControlI2VNode(KlingImage2VideoNode): +class KlingCameraControlI2VNode(comfy_io.ComfyNode): """ Kling Image to Video Camera Control Node. This node is a image to video node, but it supports controlling the camera. Duration, mode, and model_name request fields are hard-coded because camera control is only supported in pro mode with the kling-v1-5 model at 5s duration as of 2025-05-02. """ @classmethod - def INPUT_TYPES(s): - return { - "required": { - "start_frame": model_field_to_node_input( - IO.IMAGE, KlingImage2VideoRequest, "image" + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingCameraControlI2VNode", + display_name="Kling Image to Video (Camera Control)", + category="api node/video/Kling", + description="Transform still images into cinematic videos with professional camera movements that simulate real-world cinematography. Control virtual camera actions including zoom, rotation, pan, tilt, and first-person view, while maintaining focus on your original image.", + inputs=[ + comfy_io.Image.Input( + "start_frame", + tooltip="Reference Image - URL or Base64 encoded string, cannot exceed 10MB, resolution not less than 300*300px, aspect ratio between 1:2.5 ~ 2.5:1. Base64 should not include data:image prefix.", ), - "prompt": model_field_to_node_input( - IO.STRING, KlingImage2VideoRequest, "prompt", multiline=True - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - KlingImage2VideoRequest, - "negative_prompt", - multiline=True, - ), - "cfg_scale": model_field_to_node_input( - IO.FLOAT, - KlingImage2VideoRequest, - "cfg_scale", - default=0.75, - min=0.0, - max=1.0, - ), - "aspect_ratio": model_field_to_node_input( - IO.COMBO, - KlingImage2VideoRequest, + comfy_io.String.Input("prompt", multiline=True, tooltip="Positive text prompt"), + comfy_io.String.Input("negative_prompt", multiline=True, tooltip="Negative text prompt"), + comfy_io.Float.Input("cfg_scale", default=0.75, min=0.0, max=1.0), + comfy_io.Combo.Input( "aspect_ratio", - enum_type=KlingVideoGenAspectRatio, + options=[i.value for i in KlingVideoGenAspectRatio], + default="16:9", ), - "camera_control": ( - "CAMERA_CONTROL", - { - "tooltip": "Can be created using the Kling Camera Controls node. Controls the camera movement and motion during the video generation.", - }, + comfy_io.Custom("CAMERA_CONTROL").Input( + "camera_control", + tooltip="Can be created using the Kling Camera Controls node. Controls the camera movement and motion during the video generation.", ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Transform still images into cinematic videos with professional camera movements that simulate real-world cinematography. Control virtual camera actions including zoom, rotation, pan, tilt, and first-person view, while maintaining focus on your original image." - - async def api_call( - self, + @classmethod + async def execute( + cls, start_frame: torch.Tensor, prompt: str, negative_prompt: str, cfg_scale: float, aspect_ratio: str, camera_control: KlingCameraControl, - unique_id: Optional[str] = None, - **kwargs, - ): - return await super().api_call( + ) -> comfy_io.NodeOutput: + return await execute_image2video( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, model_name=KlingVideoGenModelName.kling_v1_5, start_frame=start_frame, cfg_scale=cfg_scale, - mode=KlingVideoGenMode.pro, + model_mode=KlingVideoGenMode.pro, aspect_ratio=KlingVideoGenAspectRatio(aspect_ratio), duration=KlingVideoGenDuration.field_5, prompt=prompt, negative_prompt=negative_prompt, camera_control=camera_control, - unique_id=unique_id, - **kwargs, ) -class KlingStartEndFrameNode(KlingImage2VideoNode): +class KlingStartEndFrameNode(comfy_io.ComfyNode): """ Kling First Last Frame Node. This node allows creation of a video from a first and last frame. It calls the normal image to video endpoint, but only allows the subset of input options that support the `image_tail` request field. """ - @staticmethod - def get_mode_string_mapping() -> dict[str, tuple[str, str, str]]: - """ - Returns a mapping of mode strings to their corresponding (mode, duration, model_name) tuples. - Only includes config combos that support the `image_tail` request field. - - See: [Kling API Docs Capability Map](https://app.klingai.com/global/dev/document-api/apiReference/model/skillsMap) - """ - return { - "standard mode / 5s duration / kling-v1": ("std", "5", "kling-v1"), - "pro mode / 5s duration / kling-v1": ("pro", "5", "kling-v1"), - "pro mode / 5s duration / kling-v1-5": ("pro", "5", "kling-v1-5"), - "pro mode / 10s duration / kling-v1-5": ("pro", "10", "kling-v1-5"), - "pro mode / 5s duration / kling-v1-6": ("pro", "5", "kling-v1-6"), - "pro mode / 10s duration / kling-v1-6": ("pro", "10", "kling-v1-6"), - "pro mode / 5s duration / kling-v2-1": ("pro", "5", "kling-v2-1"), - "pro mode / 10s duration / kling-v2-1": ("pro", "10", "kling-v2-1"), - } + @classmethod + def define_schema(cls) -> comfy_io.Schema: + modes = list(MODE_START_END_FRAME.keys()) + return comfy_io.Schema( + node_id="KlingStartEndFrameNode", + display_name="Kling Start-End Frame to Video", + category="api node/video/Kling", + description="Generate a video sequence that transitions between your provided start and end images. The node creates all frames in between, producing a smooth transformation from the first frame to the last.", + inputs=[ + comfy_io.Image.Input( + "start_frame", + tooltip="Reference Image - URL or Base64 encoded string, cannot exceed 10MB, resolution not less than 300*300px, aspect ratio between 1:2.5 ~ 2.5:1. Base64 should not include data:image prefix.", + ), + comfy_io.Image.Input( + "end_frame", + tooltip="Reference Image - End frame control. URL or Base64 encoded string, cannot exceed 10MB, resolution not less than 300*300px. Base64 should not include data:image prefix.", + ), + comfy_io.String.Input("prompt", multiline=True, tooltip="Positive text prompt"), + comfy_io.String.Input("negative_prompt", multiline=True, tooltip="Negative text prompt"), + comfy_io.Float.Input("cfg_scale", default=0.5, min=0.0, max=1.0), + comfy_io.Combo.Input( + "aspect_ratio", + options=[i.value for i in KlingVideoGenAspectRatio], + default="16:9", + ), + comfy_io.Combo.Input( + "mode", + options=modes, + default=modes[2], + tooltip="The configuration to use for the video generation following the format: mode / duration / model_name.", + ), + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - modes = list(KlingStartEndFrameNode.get_mode_string_mapping().keys()) - return { - "required": { - "start_frame": model_field_to_node_input( - IO.IMAGE, KlingImage2VideoRequest, "image" - ), - "end_frame": model_field_to_node_input( - IO.IMAGE, KlingImage2VideoRequest, "image_tail" - ), - "prompt": model_field_to_node_input( - IO.STRING, KlingImage2VideoRequest, "prompt", multiline=True - ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - KlingImage2VideoRequest, - "negative_prompt", - multiline=True, - ), - "cfg_scale": model_field_to_node_input( - IO.FLOAT, - KlingImage2VideoRequest, - "cfg_scale", - default=0.5, - min=0.0, - max=1.0, - ), - "aspect_ratio": model_field_to_node_input( - IO.COMBO, - KlingImage2VideoRequest, - "aspect_ratio", - enum_type=KlingVideoGenAspectRatio, - ), - "mode": ( - modes, - { - "default": modes[2], - "tooltip": "The configuration to use for the video generation following the format: mode / duration / model_name.", - }, - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - DESCRIPTION = "Generate a video sequence that transitions between your provided start and end images. The node creates all frames in between, producing a smooth transformation from the first frame to the last." - - async def api_call( - self, + async def execute( + cls, start_frame: torch.Tensor, end_frame: torch.Tensor, prompt: str, @@ -915,90 +1093,78 @@ class KlingStartEndFrameNode(KlingImage2VideoNode): cfg_scale: float, aspect_ratio: str, mode: str, - unique_id: Optional[str] = None, - **kwargs, - ): - mode, duration, model_name = KlingStartEndFrameNode.get_mode_string_mapping()[ - mode - ] - return await super().api_call( + ) -> comfy_io.NodeOutput: + mode, duration, model_name = MODE_START_END_FRAME[mode] + return await execute_image2video( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, prompt=prompt, negative_prompt=negative_prompt, model_name=model_name, start_frame=start_frame, cfg_scale=cfg_scale, - mode=mode, + model_mode=mode, aspect_ratio=aspect_ratio, duration=duration, end_frame=end_frame, - unique_id=unique_id, - **kwargs, ) -class KlingVideoExtendNode(KlingNodeBase): +class KlingVideoExtendNode(comfy_io.ComfyNode): @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": model_field_to_node_input( - IO.STRING, KlingVideoExtendRequest, "prompt", multiline=True + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingVideoExtendNode", + display_name="Kling Video Extend", + category="api node/video/Kling", + description="Kling Video Extend Node. Extend videos made by other Kling nodes. The video_id is created by using other Kling Nodes.", + inputs=[ + comfy_io.String.Input( + "prompt", + multiline=True, + tooltip="Positive text prompt for guiding the video extension", ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - KlingVideoExtendRequest, + comfy_io.String.Input( "negative_prompt", multiline=True, + tooltip="Negative text prompt for elements to avoid in the extended video", ), - "cfg_scale": model_field_to_node_input( - IO.FLOAT, - KlingVideoExtendRequest, - "cfg_scale", - default=0.5, - min=0.0, - max=1.0, + comfy_io.Float.Input("cfg_scale", default=0.5, min=0.0, max=1.0), + comfy_io.String.Input( + "video_id", + force_input=True, + tooltip="The ID of the video to be extended. Supports videos generated by text-to-video, image-to-video, and previous video extension operations. Cannot exceed 3 minutes total duration after extension.", ), - "video_id": model_field_to_node_input( - IO.STRING, KlingVideoExtendRequest, "video_id", forceInput=True - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - RETURN_TYPES = ("VIDEO", "STRING", "STRING") - RETURN_NAMES = ("VIDEO", "video_id", "duration") - DESCRIPTION = "Kling Video Extend Node. Extend videos made by other Kling nodes. The video_id is created by using other Kling Nodes." - - async def get_response( - self, task_id: str, auth_kwargs: dict[str, str], node_id: Optional[str] = None - ) -> KlingVideoExtendResponse: - return await poll_until_finished( - auth_kwargs, - ApiEndpoint( - path=f"{PATH_VIDEO_EXTEND}/{task_id}", - method=HttpMethod.GET, - request_model=EmptyRequest, - response_model=KlingVideoExtendResponse, - ), - result_url_extractor=get_video_url_from_response, - estimated_duration=AVERAGE_DURATION_VIDEO_EXTEND, - node_id=node_id, + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, ) - async def api_call( - self, + @classmethod + async def execute( + cls, prompt: str, negative_prompt: str, cfg_scale: float, video_id: str, - unique_id: Optional[str] = None, - **kwargs, - ) -> tuple[VideoFromFile, str, str]: + ) -> comfy_io.NodeOutput: validate_prompts(prompt, negative_prompt, MAX_PROMPT_LENGTH_T2V) + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_VIDEO_EXTEND, @@ -1012,560 +1178,323 @@ class KlingVideoExtendNode(KlingNodeBase): cfg_scale=cfg_scale, video_id=video_id, ), - auth_kwargs=kwargs, + auth_kwargs=auth, ) task_creation_response = await initial_operation.execute() validate_task_creation_response(task_creation_response) task_id = task_creation_response.data.task_id - final_response = await self.get_response( - task_id, auth_kwargs=kwargs, node_id=unique_id - ) - validate_video_result_response(final_response) - - video = get_video_from_response(final_response) - return await video_result_to_node_output(video) - - -class KlingVideoEffectsBase(KlingNodeBase): - """Kling Video Effects Base""" - - RETURN_TYPES = ("VIDEO", "STRING", "STRING") - RETURN_NAMES = ("VIDEO", "video_id", "duration") - - async def get_response( - self, task_id: str, auth_kwargs: dict[str, str], node_id: Optional[str] = None - ) -> KlingVideoEffectsResponse: - return await poll_until_finished( - auth_kwargs, + final_response = await poll_until_finished( + auth, ApiEndpoint( - path=f"{PATH_VIDEO_EFFECTS}/{task_id}", + path=f"{PATH_VIDEO_EXTEND}/{task_id}", method=HttpMethod.GET, request_model=EmptyRequest, - response_model=KlingVideoEffectsResponse, + response_model=KlingVideoExtendResponse, ), result_url_extractor=get_video_url_from_response, - estimated_duration=AVERAGE_DURATION_VIDEO_EFFECTS, - node_id=node_id, - ) - - async def api_call( - self, - dual_character: bool, - effect_scene: KlingDualCharacterEffectsScene | KlingSingleImageEffectsScene, - model_name: str, - duration: KlingVideoGenDuration, - image_1: torch.Tensor, - image_2: Optional[torch.Tensor] = None, - mode: Optional[KlingVideoGenMode] = None, - unique_id: Optional[str] = None, - **kwargs, - ): - if dual_character: - request_input_field = KlingDualCharacterEffectInput( - model_name=model_name, - mode=mode, - images=[ - tensor_to_base64_string(image_1), - tensor_to_base64_string(image_2), - ], - duration=duration, - ) - else: - request_input_field = KlingSingleImageEffectInput( - model_name=model_name, - image=tensor_to_base64_string(image_1), - duration=duration, - ) - - initial_operation = SynchronousOperation( - endpoint=ApiEndpoint( - path=PATH_VIDEO_EFFECTS, - method=HttpMethod.POST, - request_model=KlingVideoEffectsRequest, - response_model=KlingVideoEffectsResponse, - ), - request=KlingVideoEffectsRequest( - effect_scene=effect_scene, - input=request_input_field, - ), - auth_kwargs=kwargs, - ) - - task_creation_response = await initial_operation.execute() - validate_task_creation_response(task_creation_response) - task_id = task_creation_response.data.task_id - - final_response = await self.get_response( - task_id, auth_kwargs=kwargs, node_id=unique_id + estimated_duration=AVERAGE_DURATION_VIDEO_EXTEND, + node_id=cls.hidden.unique_id, ) validate_video_result_response(final_response) video = get_video_from_response(final_response) - return await video_result_to_node_output(video) + return comfy_io.NodeOutput(await download_url_to_video_output(str(video.url)), str(video.id), str(video.duration)) -class KlingDualCharacterVideoEffectNode(KlingVideoEffectsBase): +class KlingDualCharacterVideoEffectNode(comfy_io.ComfyNode): """Kling Dual Character Video Effect Node""" @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image_left": (IO.IMAGE, {"tooltip": "Left side image"}), - "image_right": (IO.IMAGE, {"tooltip": "Right side image"}), - "effect_scene": model_field_to_node_input( - IO.COMBO, - KlingVideoEffectsRequest, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingDualCharacterVideoEffectNode", + display_name="Kling Dual Character Video Effects", + category="api node/video/Kling", + description="Achieve different special effects when generating a video based on the effect_scene. First image will be positioned on left side, second on right side of the composite.", + inputs=[ + comfy_io.Image.Input("image_left", tooltip="Left side image"), + comfy_io.Image.Input("image_right", tooltip="Right side image"), + comfy_io.Combo.Input( "effect_scene", - enum_type=KlingDualCharacterEffectsScene, + options=[i.value for i in KlingDualCharacterEffectsScene], ), - "model_name": model_field_to_node_input( - IO.COMBO, - KlingDualCharacterEffectInput, + comfy_io.Combo.Input( "model_name", - enum_type=KlingCharacterEffectModelName, + options=[i.value for i in KlingCharacterEffectModelName], + default="kling-v1", ), - "mode": model_field_to_node_input( - IO.COMBO, - KlingDualCharacterEffectInput, + comfy_io.Combo.Input( "mode", - enum_type=KlingVideoGenMode, + options=[i.value for i in KlingVideoGenMode], + default="std", ), - "duration": model_field_to_node_input( - IO.COMBO, - KlingDualCharacterEffectInput, + comfy_io.Combo.Input( "duration", - enum_type=KlingVideoGenDuration, + options=[i.value for i in KlingVideoGenDuration], ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Achieve different special effects when generating a video based on the effect_scene. First image will be positioned on left side, second on right side of the composite." - RETURN_TYPES = ("VIDEO", "STRING") - RETURN_NAMES = ("VIDEO", "duration") - - async def api_call( - self, + @classmethod + async def execute( + cls, image_left: torch.Tensor, image_right: torch.Tensor, effect_scene: KlingDualCharacterEffectsScene, model_name: KlingCharacterEffectModelName, mode: KlingVideoGenMode, duration: KlingVideoGenDuration, - unique_id: Optional[str] = None, - **kwargs, - ): - video, _, duration = await super().api_call( + ) -> comfy_io.NodeOutput: + video, _, duration = await execute_video_effect( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, dual_character=True, effect_scene=effect_scene, model_name=model_name, - mode=mode, + model_mode=mode, duration=duration, image_1=image_left, image_2=image_right, - unique_id=unique_id, - **kwargs, ) return video, duration -class KlingSingleImageVideoEffectNode(KlingVideoEffectsBase): +class KlingSingleImageVideoEffectNode(comfy_io.ComfyNode): """Kling Single Image Video Effect Node""" @classmethod - def INPUT_TYPES(s): - return { - "required": { - "image": ( - IO.IMAGE, - { - "tooltip": " Reference Image. URL or Base64 encoded string (without data:image prefix). File size cannot exceed 10MB, resolution not less than 300*300px, aspect ratio between 1:2.5 ~ 2.5:1" - }, - ), - "effect_scene": model_field_to_node_input( - IO.COMBO, - KlingVideoEffectsRequest, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingSingleImageVideoEffectNode", + display_name="Kling Video Effects", + category="api node/video/Kling", + description="Achieve different special effects when generating a video based on the effect_scene.", + inputs=[ + comfy_io.Image.Input("image", tooltip=" Reference Image. URL or Base64 encoded string (without data:image prefix). File size cannot exceed 10MB, resolution not less than 300*300px, aspect ratio between 1:2.5 ~ 2.5:1"), + comfy_io.Combo.Input( "effect_scene", - enum_type=KlingSingleImageEffectsScene, + options=[i.value for i in KlingSingleImageEffectsScene], ), - "model_name": model_field_to_node_input( - IO.COMBO, - KlingSingleImageEffectInput, + comfy_io.Combo.Input( "model_name", - enum_type=KlingSingleImageEffectModelName, + options=[i.value for i in KlingSingleImageEffectModelName], ), - "duration": model_field_to_node_input( - IO.COMBO, - KlingSingleImageEffectInput, + comfy_io.Combo.Input( "duration", - enum_type=KlingVideoGenDuration, + options=[i.value for i in KlingVideoGenDuration], ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Achieve different special effects when generating a video based on the effect_scene." - - async def api_call( - self, + @classmethod + async def execute( + cls, image: torch.Tensor, effect_scene: KlingSingleImageEffectsScene, model_name: KlingSingleImageEffectModelName, duration: KlingVideoGenDuration, - unique_id: Optional[str] = None, - **kwargs, - ): - return await super().api_call( + ) -> comfy_io.NodeOutput: + return await execute_video_effect( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, dual_character=False, effect_scene=effect_scene, model_name=model_name, duration=duration, image_1=image, - unique_id=unique_id, - **kwargs, ) -class KlingLipSyncBase(KlingNodeBase): - """Kling Lip Sync Base""" - - RETURN_TYPES = ("VIDEO", "STRING", "STRING") - RETURN_NAMES = ("VIDEO", "video_id", "duration") - - def validate_lip_sync_video(self, video: VideoInput): - """ - Validates the input video adheres to the expectations of the Kling Lip Sync API: - - Video length does not exceed 10s and is not shorter than 2s - - Length and width dimensions should both be between 720px and 1920px - - See: https://app.klingai.com/global/dev/document-api/apiReference/model/videoTolip - """ - validate_video_dimensions(video, 720, 1920) - validate_video_duration(video, 2, 10) - - def validate_text(self, text: str): - if not text: - raise ValueError("Text is required") - if len(text) > MAX_PROMPT_LENGTH_LIP_SYNC: - raise ValueError( - f"Text is too long. Maximum length is {MAX_PROMPT_LENGTH_LIP_SYNC} characters." - ) - - async def get_response( - self, task_id: str, auth_kwargs: dict[str, str], node_id: Optional[str] = None - ) -> KlingLipSyncResponse: - """Polls the Kling API endpoint until the task reaches a terminal state.""" - return await poll_until_finished( - auth_kwargs, - ApiEndpoint( - path=f"{PATH_LIP_SYNC}/{task_id}", - method=HttpMethod.GET, - request_model=EmptyRequest, - response_model=KlingLipSyncResponse, - ), - result_url_extractor=get_video_url_from_response, - estimated_duration=AVERAGE_DURATION_LIP_SYNC, - node_id=node_id, - ) - - async def api_call( - self, - video: VideoInput, - audio: Optional[AudioInput] = None, - voice_language: Optional[str] = None, - mode: Optional[str] = None, - text: Optional[str] = None, - voice_speed: Optional[float] = None, - voice_id: Optional[str] = None, - unique_id: Optional[str] = None, - **kwargs, - ) -> tuple[VideoFromFile, str, str]: - if text: - self.validate_text(text) - self.validate_lip_sync_video(video) - - # Upload video to Comfy API and get download URL - video_url = await upload_video_to_comfyapi(video, auth_kwargs=kwargs) - logging.info("Uploaded video to Comfy API. URL: %s", video_url) - - # Upload the audio file to Comfy API and get download URL - if audio: - audio_url = await upload_audio_to_comfyapi(audio, auth_kwargs=kwargs) - logging.info("Uploaded audio to Comfy API. URL: %s", audio_url) - else: - audio_url = None - - initial_operation = SynchronousOperation( - endpoint=ApiEndpoint( - path=PATH_LIP_SYNC, - method=HttpMethod.POST, - request_model=KlingLipSyncRequest, - response_model=KlingLipSyncResponse, - ), - request=KlingLipSyncRequest( - input=KlingLipSyncInputObject( - video_url=video_url, - mode=mode, - text=text, - voice_language=voice_language, - voice_speed=voice_speed, - audio_type="url", - audio_url=audio_url, - voice_id=voice_id, - ), - ), - auth_kwargs=kwargs, - ) - - task_creation_response = await initial_operation.execute() - validate_task_creation_response(task_creation_response) - task_id = task_creation_response.data.task_id - - final_response = await self.get_response( - task_id, auth_kwargs=kwargs, node_id=unique_id - ) - validate_video_result_response(final_response) - - video = get_video_from_response(final_response) - return await video_result_to_node_output(video) - - -class KlingLipSyncAudioToVideoNode(KlingLipSyncBase): +class KlingLipSyncAudioToVideoNode(comfy_io.ComfyNode): """Kling Lip Sync Audio to Video Node. Syncs mouth movements in a video file to the audio content of an audio file.""" @classmethod - def INPUT_TYPES(s): - return { - "required": { - "video": (IO.VIDEO, {}), - "audio": (IO.AUDIO, {}), - "voice_language": model_field_to_node_input( - IO.COMBO, - KlingLipSyncInputObject, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingLipSyncAudioToVideoNode", + display_name="Kling Lip Sync Video with Audio", + category="api node/video/Kling", + description="Kling Lip Sync Audio to Video Node. Syncs mouth movements in a video file to the audio content of an audio file. When using, ensure that the audio contains clearly distinguishable vocals and that the video contains a distinct face. The audio file should not be larger than 5MB. The video file should not be larger than 100MB, should have height/width between 720px and 1920px, and should be between 2s and 10s in length.", + inputs=[ + comfy_io.Video.Input("video"), + comfy_io.Audio.Input("audio"), + comfy_io.Combo.Input( "voice_language", - enum_type=KlingLipSyncVoiceLanguage, + options=[i.value for i in KlingLipSyncVoiceLanguage], + default="en", ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) - DESCRIPTION = "Kling Lip Sync Audio to Video Node. Syncs mouth movements in a video file to the audio content of an audio file. When using, ensure that the audio contains clearly distinguishable vocals and that the video contains a distinct face. The audio file should not be larger than 5MB. The video file should not be larger than 100MB, should have height/width between 720px and 1920px, and should be between 2s and 10s in length." - - async def api_call( - self, + @classmethod + async def execute( + cls, video: VideoInput, audio: AudioInput, voice_language: str, - unique_id: Optional[str] = None, - **kwargs, - ): - return await super().api_call( + ) -> comfy_io.NodeOutput: + return await execute_lipsync( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, video=video, audio=audio, voice_language=voice_language, - mode="audio2video", - unique_id=unique_id, - **kwargs, + model_mode="audio2video", ) -class KlingLipSyncTextToVideoNode(KlingLipSyncBase): +class KlingLipSyncTextToVideoNode(comfy_io.ComfyNode): """Kling Lip Sync Text to Video Node. Syncs mouth movements in a video file to a text prompt.""" - @staticmethod - def get_voice_config() -> dict[str, tuple[str, str]]: - return { - # English voices - "Melody": ("girlfriend_4_speech02", "en"), - "Sunny": ("genshin_vindi2", "en"), - "Sage": ("zhinen_xuesheng", "en"), - "Ace": ("AOT", "en"), - "Blossom": ("ai_shatang", "en"), - "Peppy": ("genshin_klee2", "en"), - "Dove": ("genshin_kirara", "en"), - "Shine": ("ai_kaiya", "en"), - "Anchor": ("oversea_male1", "en"), - "Lyric": ("ai_chenjiahao_712", "en"), - "Tender": ("chat1_female_new-3", "en"), - "Siren": ("chat_0407_5-1", "en"), - "Zippy": ("cartoon-boy-07", "en"), - "Bud": ("uk_boy1", "en"), - "Sprite": ("cartoon-girl-01", "en"), - "Candy": ("PeppaPig_platform", "en"), - "Beacon": ("ai_huangzhong_712", "en"), - "Rock": ("ai_huangyaoshi_712", "en"), - "Titan": ("ai_laoguowang_712", "en"), - "Grace": ("chengshu_jiejie", "en"), - "Helen": ("you_pingjing", "en"), - "Lore": ("calm_story1", "en"), - "Crag": ("uk_man2", "en"), - "Prattle": ("laopopo_speech02", "en"), - "Hearth": ("heainainai_speech02", "en"), - "The Reader": ("reader_en_m-v1", "en"), - "Commercial Lady": ("commercial_lady_en_f-v1", "en"), - # Chinese voices - "阳光少年": ("genshin_vindi2", "zh"), - "懂事小弟": ("zhinen_xuesheng", "zh"), - "运动少年": ("tiyuxi_xuedi", "zh"), - "青春少女": ("ai_shatang", "zh"), - "温柔小妹": ("genshin_klee2", "zh"), - "元气少女": ("genshin_kirara", "zh"), - "阳光男生": ("ai_kaiya", "zh"), - "幽默小哥": ("tiexin_nanyou", "zh"), - "文艺小哥": ("ai_chenjiahao_712", "zh"), - "甜美邻家": ("girlfriend_1_speech02", "zh"), - "温柔姐姐": ("chat1_female_new-3", "zh"), - "职场女青": ("girlfriend_2_speech02", "zh"), - "活泼男童": ("cartoon-boy-07", "zh"), - "俏皮女童": ("cartoon-girl-01", "zh"), - "稳重老爸": ("ai_huangyaoshi_712", "zh"), - "温柔妈妈": ("you_pingjing", "zh"), - "严肃上司": ("ai_laoguowang_712", "zh"), - "优雅贵妇": ("chengshu_jiejie", "zh"), - "慈祥爷爷": ("zhuxi_speech02", "zh"), - "唠叨爷爷": ("uk_oldman3", "zh"), - "唠叨奶奶": ("laopopo_speech02", "zh"), - "和蔼奶奶": ("heainainai_speech02", "zh"), - "东北老铁": ("dongbeilaotie_speech02", "zh"), - "重庆小伙": ("chongqingxiaohuo_speech02", "zh"), - "四川妹子": ("chuanmeizi_speech02", "zh"), - "潮汕大叔": ("chaoshandashu_speech02", "zh"), - "台湾男生": ("ai_taiwan_man2_speech02", "zh"), - "西安掌柜": ("xianzhanggui_speech02", "zh"), - "天津姐姐": ("tianjinjiejie_speech02", "zh"), - "新闻播报男": ("diyinnansang_DB_CN_M_04-v2", "zh"), - "译制片男": ("yizhipiannan-v1", "zh"), - "撒娇女友": ("tianmeixuemei-v1", "zh"), - "刀片烟嗓": ("daopianyansang-v1", "zh"), - "乖巧正太": ("mengwa-v1", "zh"), - } + @classmethod + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingLipSyncTextToVideoNode", + display_name="Kling Lip Sync Video with Text", + category="api node/video/Kling", + description="Kling Lip Sync Text to Video Node. Syncs mouth movements in a video file to a text prompt. The video file should not be larger than 100MB, should have height/width between 720px and 1920px, and should be between 2s and 10s in length.", + inputs=[ + comfy_io.Video.Input("video"), + comfy_io.String.Input( + "text", + multiline=True, + tooltip="Text Content for Lip-Sync Video Generation. Required when mode is text2video. Maximum length is 120 characters.", + ), + comfy_io.Combo.Input( + "voice", + options=list(VOICES_CONFIG.keys()), + default="Melody", + ), + comfy_io.Float.Input( + "voice_speed", + default=1, + min=0.8, + max=2.0, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Speech Rate. Valid range: 0.8~2.0, accurate to one decimal place.", + ), + ], + outputs=[ + comfy_io.Video.Output(), + comfy_io.String.Output(display_name="video_id"), + comfy_io.String.Output(display_name="duration"), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) @classmethod - def INPUT_TYPES(s): - voice_options = list(s.get_voice_config().keys()) - return { - "required": { - "video": (IO.VIDEO, {}), - "text": model_field_to_node_input( - IO.STRING, KlingLipSyncInputObject, "text", multiline=True - ), - "voice": (voice_options, {"default": voice_options[0]}), - "voice_speed": model_field_to_node_input( - IO.FLOAT, KlingLipSyncInputObject, "voice_speed", slider=True - ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - DESCRIPTION = "Kling Lip Sync Text to Video Node. Syncs mouth movements in a video file to a text prompt. The video file should not be larger than 100MB, should have height/width between 720px and 1920px, and should be between 2s and 10s in length." - - async def api_call( - self, + async def execute( + cls, video: VideoInput, text: str, voice: str, voice_speed: float, - unique_id: Optional[str] = None, - **kwargs, - ): - voice_id, voice_language = KlingLipSyncTextToVideoNode.get_voice_config()[voice] - return await super().api_call( + ) -> comfy_io.NodeOutput: + voice_id, voice_language = VOICES_CONFIG[voice] + return await execute_lipsync( + auth_kwargs={ + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + }, + node_id=cls.hidden.unique_id, video=video, text=text, voice_language=voice_language, voice_id=voice_id, voice_speed=voice_speed, - mode="text2video", - unique_id=unique_id, - **kwargs, + model_mode="text2video", ) -class KlingImageGenerationBase(KlingNodeBase): - """Kling Image Generation Base Node.""" - - RETURN_TYPES = ("IMAGE",) - CATEGORY = "api node/image/Kling" - - def validate_prompt(self, prompt: str, negative_prompt: Optional[str] = None): - if not prompt or len(prompt) > MAX_PROMPT_LENGTH_IMAGE_GEN: - raise ValueError( - f"Prompt must be less than {MAX_PROMPT_LENGTH_IMAGE_GEN} characters" - ) - if negative_prompt and len(negative_prompt) > MAX_PROMPT_LENGTH_IMAGE_GEN: - raise ValueError( - f"Negative prompt must be less than {MAX_PROMPT_LENGTH_IMAGE_GEN} characters" - ) - - -class KlingVirtualTryOnNode(KlingImageGenerationBase): +class KlingVirtualTryOnNode(comfy_io.ComfyNode): """Kling Virtual Try On Node.""" @classmethod - def INPUT_TYPES(s): - return { - "required": { - "human_image": (IO.IMAGE, {}), - "cloth_image": (IO.IMAGE, {}), - "model_name": model_field_to_node_input( - IO.COMBO, - KlingVirtualTryOnRequest, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingVirtualTryOnNode", + display_name="Kling Virtual Try On", + category="api node/image/Kling", + description="Kling Virtual Try On Node. Input a human image and a cloth image to try on the cloth on the human. You can merge multiple clothing item pictures into one image with a white background.", + inputs=[ + comfy_io.Image.Input("human_image"), + comfy_io.Image.Input("cloth_image"), + comfy_io.Combo.Input( "model_name", - enum_type=KlingVirtualTryOnModelName, + options=[i.value for i in KlingVirtualTryOnModelName], + default="kolors-virtual-try-on-v1", ), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - DESCRIPTION = "Kling Virtual Try On Node. Input a human image and a cloth image to try on the cloth on the human. You can merge multiple clothing item pictures into one image with a white background." - - async def get_response( - self, task_id: str, auth_kwargs: dict[str, str], node_id: Optional[str] = None - ) -> KlingVirtualTryOnResponse: - return await poll_until_finished( - auth_kwargs, - ApiEndpoint( - path=f"{PATH_VIRTUAL_TRY_ON}/{task_id}", - method=HttpMethod.GET, - request_model=EmptyRequest, - response_model=KlingVirtualTryOnResponse, - ), - result_url_extractor=get_images_urls_from_response, - estimated_duration=AVERAGE_DURATION_VIRTUAL_TRY_ON, - node_id=node_id, + ], + outputs=[ + comfy_io.Image.Output(), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, ) - async def api_call( - self, + @classmethod + async def execute( + cls, human_image: torch.Tensor, cloth_image: torch.Tensor, model_name: KlingVirtualTryOnModelName, - unique_id: Optional[str] = None, - **kwargs, - ): + ) -> comfy_io.NodeOutput: + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_VIRTUAL_TRY_ON, @@ -1578,113 +1507,99 @@ class KlingVirtualTryOnNode(KlingImageGenerationBase): cloth_image=tensor_to_base64_string(cloth_image), model_name=model_name, ), - auth_kwargs=kwargs, + auth_kwargs=auth, ) task_creation_response = await initial_operation.execute() validate_task_creation_response(task_creation_response) task_id = task_creation_response.data.task_id - final_response = await self.get_response( - task_id, auth_kwargs=kwargs, node_id=unique_id + final_response = await poll_until_finished( + auth, + ApiEndpoint( + path=f"{PATH_VIRTUAL_TRY_ON}/{task_id}", + method=HttpMethod.GET, + request_model=EmptyRequest, + response_model=KlingVirtualTryOnResponse, + ), + result_url_extractor=get_images_urls_from_response, + estimated_duration=AVERAGE_DURATION_VIRTUAL_TRY_ON, + node_id=cls.hidden.unique_id, ) validate_image_result_response(final_response) images = get_images_from_response(final_response) - return (await image_result_to_node_output(images),) + return comfy_io.NodeOutput(await image_result_to_node_output(images)) -class KlingImageGenerationNode(KlingImageGenerationBase): +class KlingImageGenerationNode(comfy_io.ComfyNode): """Kling Image Generation Node. Generate an image from a text prompt with an optional reference image.""" @classmethod - def INPUT_TYPES(s): - return { - "required": { - "prompt": model_field_to_node_input( - IO.STRING, - KlingImageGenerationsRequest, - "prompt", - multiline=True, - max_length=MAX_PROMPT_LENGTH_IMAGE_GEN, + def define_schema(cls) -> comfy_io.Schema: + return comfy_io.Schema( + node_id="KlingImageGenerationNode", + display_name="Kling Image Generation", + category="api node/image/Kling", + description="Kling Image Generation Node. Generate an image from a text prompt with an optional reference image.", + inputs=[ + comfy_io.String.Input("prompt", multiline=True, tooltip="Positive text prompt"), + comfy_io.String.Input("negative_prompt", multiline=True, tooltip="Negative text prompt"), + comfy_io.Combo.Input( + "image_type", + options=[i.value for i in KlingImageGenImageReferenceType], ), - "negative_prompt": model_field_to_node_input( - IO.STRING, - KlingImageGenerationsRequest, - "negative_prompt", - multiline=True, - ), - "image_type": model_field_to_node_input( - IO.COMBO, - KlingImageGenerationsRequest, - "image_reference", - enum_type=KlingImageGenImageReferenceType, - ), - "image_fidelity": model_field_to_node_input( - IO.FLOAT, - KlingImageGenerationsRequest, + comfy_io.Float.Input( "image_fidelity", - slider=True, + default=0.5, + min=0.0, + max=1.0, step=0.01, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Reference intensity for user-uploaded images", ), - "human_fidelity": model_field_to_node_input( - IO.FLOAT, - KlingImageGenerationsRequest, + comfy_io.Float.Input( "human_fidelity", - slider=True, + default=0.45, + min=0.0, + max=1.0, step=0.01, + display_mode=comfy_io.NumberDisplay.slider, + tooltip="Subject reference similarity", ), - "model_name": model_field_to_node_input( - IO.COMBO, - KlingImageGenerationsRequest, + comfy_io.Combo.Input( "model_name", - enum_type=KlingImageGenModelName, + options=[i.value for i in KlingImageGenModelName], + default="kling-v1", ), - "aspect_ratio": model_field_to_node_input( - IO.COMBO, - KlingImageGenerationsRequest, + comfy_io.Combo.Input( "aspect_ratio", - enum_type=KlingImageGenAspectRatio, + options=[i.value for i in KlingImageGenAspectRatio], + default="16:9", ), - "n": model_field_to_node_input( - IO.INT, - KlingImageGenerationsRequest, + comfy_io.Int.Input( "n", + default=1, + min=1, + max=9, + tooltip="Number of generated images", ), - }, - "optional": { - "image": (IO.IMAGE, {}), - }, - "hidden": { - "auth_token": "AUTH_TOKEN_COMFY_ORG", - "comfy_api_key": "API_KEY_COMFY_ORG", - "unique_id": "UNIQUE_ID", - }, - } - - DESCRIPTION = "Kling Image Generation Node. Generate an image from a text prompt with an optional reference image." - - async def get_response( - self, - task_id: str, - auth_kwargs: Optional[dict[str, str]], - node_id: Optional[str] = None, - ) -> KlingImageGenerationsResponse: - return await poll_until_finished( - auth_kwargs, - ApiEndpoint( - path=f"{PATH_IMAGE_GENERATIONS}/{task_id}", - method=HttpMethod.GET, - request_model=EmptyRequest, - response_model=KlingImageGenerationsResponse, - ), - result_url_extractor=get_images_urls_from_response, - estimated_duration=AVERAGE_DURATION_IMAGE_GEN, - node_id=node_id, + comfy_io.Image.Input("image", optional=True), + ], + outputs=[ + comfy_io.Image.Output(), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, ) - async def api_call( - self, + @classmethod + async def execute( + cls, model_name: KlingImageGenModelName, prompt: str, negative_prompt: str, @@ -1694,10 +1609,9 @@ class KlingImageGenerationNode(KlingImageGenerationBase): n: int, aspect_ratio: KlingImageGenAspectRatio, image: Optional[torch.Tensor] = None, - unique_id: Optional[str] = None, - **kwargs, - ): - self.validate_prompt(prompt, negative_prompt) + ) -> comfy_io.NodeOutput: + validate_string(prompt, field_name="prompt", min_length=1, max_length=MAX_PROMPT_LENGTH_IMAGE_GEN) + validate_string(negative_prompt, field_name="negative_prompt", max_length=MAX_PROMPT_LENGTH_IMAGE_GEN) if image is None: image_type = None @@ -1706,6 +1620,10 @@ class KlingImageGenerationNode(KlingImageGenerationBase): else: image = tensor_to_base64_string(image) + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } initial_operation = SynchronousOperation( endpoint=ApiEndpoint( path=PATH_IMAGE_GENERATIONS, @@ -1724,50 +1642,50 @@ class KlingImageGenerationNode(KlingImageGenerationBase): n=n, aspect_ratio=aspect_ratio, ), - auth_kwargs=kwargs, + auth_kwargs=auth, ) task_creation_response = await initial_operation.execute() validate_task_creation_response(task_creation_response) task_id = task_creation_response.data.task_id - final_response = await self.get_response( - task_id, auth_kwargs=kwargs, node_id=unique_id + final_response = await poll_until_finished( + auth, + ApiEndpoint( + path=f"{PATH_IMAGE_GENERATIONS}/{task_id}", + method=HttpMethod.GET, + request_model=EmptyRequest, + response_model=KlingImageGenerationsResponse, + ), + result_url_extractor=get_images_urls_from_response, + estimated_duration=AVERAGE_DURATION_IMAGE_GEN, + node_id=cls.hidden.unique_id, ) validate_image_result_response(final_response) images = get_images_from_response(final_response) - return (await image_result_to_node_output(images),) + return comfy_io.NodeOutput(await image_result_to_node_output(images)) -NODE_CLASS_MAPPINGS = { - "KlingCameraControls": KlingCameraControls, - "KlingTextToVideoNode": KlingTextToVideoNode, - "KlingImage2VideoNode": KlingImage2VideoNode, - "KlingCameraControlI2VNode": KlingCameraControlI2VNode, - "KlingCameraControlT2VNode": KlingCameraControlT2VNode, - "KlingStartEndFrameNode": KlingStartEndFrameNode, - "KlingVideoExtendNode": KlingVideoExtendNode, - "KlingLipSyncAudioToVideoNode": KlingLipSyncAudioToVideoNode, - "KlingLipSyncTextToVideoNode": KlingLipSyncTextToVideoNode, - "KlingVirtualTryOnNode": KlingVirtualTryOnNode, - "KlingImageGenerationNode": KlingImageGenerationNode, - "KlingSingleImageVideoEffectNode": KlingSingleImageVideoEffectNode, - "KlingDualCharacterVideoEffectNode": KlingDualCharacterVideoEffectNode, -} +class KlingExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: + return [ + KlingCameraControls, + KlingTextToVideoNode, + KlingImage2VideoNode, + KlingCameraControlI2VNode, + KlingCameraControlT2VNode, + KlingStartEndFrameNode, + KlingVideoExtendNode, + KlingLipSyncAudioToVideoNode, + KlingLipSyncTextToVideoNode, + KlingVirtualTryOnNode, + KlingImageGenerationNode, + KlingSingleImageVideoEffectNode, + KlingDualCharacterVideoEffectNode, + ] -NODE_DISPLAY_NAME_MAPPINGS = { - "KlingCameraControls": "Kling Camera Controls", - "KlingTextToVideoNode": "Kling Text to Video", - "KlingImage2VideoNode": "Kling Image to Video", - "KlingCameraControlI2VNode": "Kling Image to Video (Camera Control)", - "KlingCameraControlT2VNode": "Kling Text to Video (Camera Control)", - "KlingStartEndFrameNode": "Kling Start-End Frame to Video", - "KlingVideoExtendNode": "Kling Video Extend", - "KlingLipSyncAudioToVideoNode": "Kling Lip Sync Video with Audio", - "KlingLipSyncTextToVideoNode": "Kling Lip Sync Video with Text", - "KlingVirtualTryOnNode": "Kling Virtual Try On", - "KlingImageGenerationNode": "Kling Image Generation", - "KlingSingleImageVideoEffectNode": "Kling Video Effects", - "KlingDualCharacterVideoEffectNode": "Kling Dual Character Video Effects", -} + +async def comfy_entrypoint() -> KlingExtension: + return KlingExtension() From 8aea746212dc1bb1601b4dc5e8c8093d2221d89c Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Mon, 6 Oct 2025 19:08:08 -0700 Subject: [PATCH 107/114] Implement gemma 3 as a text encoder. (#10241) Not useful yet. --- comfy/model_detection.py | 4 +- comfy/sd.py | 7 ++ comfy/text_encoders/llama.py | 133 +++++++++++++++++++++++++++------ comfy/text_encoders/lumina2.py | 26 ++++++- 4 files changed, 142 insertions(+), 28 deletions(-) diff --git a/comfy/model_detection.py b/comfy/model_detection.py index 46415c17a..7677617c0 100644 --- a/comfy/model_detection.py +++ b/comfy/model_detection.py @@ -365,8 +365,8 @@ def detect_unet_config(state_dict, key_prefix, metadata=None): dit_config["patch_size"] = 2 dit_config["in_channels"] = 16 dit_config["dim"] = 2304 - dit_config["cap_feat_dim"] = 2304 - dit_config["n_layers"] = 26 + dit_config["cap_feat_dim"] = state_dict['{}cap_embedder.1.weight'.format(key_prefix)].shape[1] + dit_config["n_layers"] = count_blocks(state_dict_keys, '{}layers.'.format(key_prefix) + '{}.') dit_config["n_heads"] = 24 dit_config["n_kv_heads"] = 8 dit_config["qk_norm"] = True diff --git a/comfy/sd.py b/comfy/sd.py index be225ad03..f2d95f85a 100644 --- a/comfy/sd.py +++ b/comfy/sd.py @@ -890,6 +890,7 @@ class TEModel(Enum): QWEN25_3B = 10 QWEN25_7B = 11 BYT5_SMALL_GLYPH = 12 + GEMMA_3_4B = 13 def detect_te_model(sd): if "text_model.encoder.layers.30.mlp.fc1.weight" in sd: @@ -912,6 +913,8 @@ def detect_te_model(sd): return TEModel.BYT5_SMALL_GLYPH return TEModel.T5_BASE if 'model.layers.0.post_feedforward_layernorm.weight' in sd: + if 'model.layers.0.self_attn.q_norm.weight' in sd: + return TEModel.GEMMA_3_4B return TEModel.GEMMA_2_2B if 'model.layers.0.self_attn.k_proj.bias' in sd: weight = sd['model.layers.0.self_attn.k_proj.bias'] @@ -1016,6 +1019,10 @@ def load_text_encoder_state_dicts(state_dicts=[], embedding_directory=None, clip clip_target.clip = comfy.text_encoders.lumina2.te(**llama_detect(clip_data)) clip_target.tokenizer = comfy.text_encoders.lumina2.LuminaTokenizer tokenizer_data["spiece_model"] = clip_data[0].get("spiece_model", None) + elif te_model == TEModel.GEMMA_3_4B: + clip_target.clip = comfy.text_encoders.lumina2.te(**llama_detect(clip_data), model_type="gemma3_4b") + clip_target.tokenizer = comfy.text_encoders.lumina2.NTokenizer + tokenizer_data["spiece_model"] = clip_data[0].get("spiece_model", None) elif te_model == TEModel.LLAMA3_8: clip_target.clip = comfy.text_encoders.hidream.hidream_clip(**llama_detect(clip_data), clip_l=False, clip_g=False, t5=False, llama=True, dtype_t5=None, t5xxl_scaled_fp8=None) diff --git a/comfy/text_encoders/llama.py b/comfy/text_encoders/llama.py index c5a48ba9f..c050759fe 100644 --- a/comfy/text_encoders/llama.py +++ b/comfy/text_encoders/llama.py @@ -3,6 +3,7 @@ import torch.nn as nn from dataclasses import dataclass from typing import Optional, Any import math +import logging from comfy.ldm.modules.attention import optimized_attention_for_device import comfy.model_management @@ -28,6 +29,9 @@ class Llama2Config: mlp_activation = "silu" qkv_bias = False rope_dims = None + q_norm = None + k_norm = None + rope_scale = None @dataclass class Qwen25_3BConfig: @@ -46,6 +50,9 @@ class Qwen25_3BConfig: mlp_activation = "silu" qkv_bias = True rope_dims = None + q_norm = None + k_norm = None + rope_scale = None @dataclass class Qwen25_7BVLI_Config: @@ -64,6 +71,9 @@ class Qwen25_7BVLI_Config: mlp_activation = "silu" qkv_bias = True rope_dims = [16, 24, 24] + q_norm = None + k_norm = None + rope_scale = None @dataclass class Gemma2_2B_Config: @@ -82,6 +92,32 @@ class Gemma2_2B_Config: mlp_activation = "gelu_pytorch_tanh" qkv_bias = False rope_dims = None + q_norm = None + k_norm = None + sliding_attention = None + rope_scale = None + +@dataclass +class Gemma3_4B_Config: + vocab_size: int = 262208 + hidden_size: int = 2560 + intermediate_size: int = 10240 + num_hidden_layers: int = 34 + num_attention_heads: int = 8 + num_key_value_heads: int = 4 + max_position_embeddings: int = 131072 + rms_norm_eps: float = 1e-6 + rope_theta = [10000.0, 1000000.0] + transformer_type: str = "gemma3" + head_dim = 256 + rms_norm_add = True + mlp_activation = "gelu_pytorch_tanh" + qkv_bias = False + rope_dims = None + q_norm = "gemma3" + k_norm = "gemma3" + sliding_attention = [False, False, False, False, False, 1024] + rope_scale = [1.0, 8.0] class RMSNorm(nn.Module): def __init__(self, dim: int, eps: float = 1e-5, add=False, device=None, dtype=None): @@ -106,25 +142,40 @@ def rotate_half(x): return torch.cat((-x2, x1), dim=-1) -def precompute_freqs_cis(head_dim, position_ids, theta, rope_dims=None, device=None): - theta_numerator = torch.arange(0, head_dim, 2, device=device).float() - inv_freq = 1.0 / (theta ** (theta_numerator / head_dim)) +def precompute_freqs_cis(head_dim, position_ids, theta, rope_scale=None, rope_dims=None, device=None): + if not isinstance(theta, list): + theta = [theta] - inv_freq_expanded = inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1) - position_ids_expanded = position_ids[:, None, :].float() - freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(1, 2) - emb = torch.cat((freqs, freqs), dim=-1) - cos = emb.cos() - sin = emb.sin() - if rope_dims is not None and position_ids.shape[0] > 1: - mrope_section = rope_dims * 2 - cos = torch.cat([m[i % 3] for i, m in enumerate(cos.split(mrope_section, dim=-1))], dim=-1).unsqueeze(0) - sin = torch.cat([m[i % 3] for i, m in enumerate(sin.split(mrope_section, dim=-1))], dim=-1).unsqueeze(0) - else: - cos = cos.unsqueeze(1) - sin = sin.unsqueeze(1) + out = [] + for index, t in enumerate(theta): + theta_numerator = torch.arange(0, head_dim, 2, device=device).float() + inv_freq = 1.0 / (t ** (theta_numerator / head_dim)) - return (cos, sin) + if rope_scale is not None: + if isinstance(rope_scale, list): + inv_freq /= rope_scale[index] + else: + inv_freq /= rope_scale + + inv_freq_expanded = inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1) + position_ids_expanded = position_ids[:, None, :].float() + freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(1, 2) + emb = torch.cat((freqs, freqs), dim=-1) + cos = emb.cos() + sin = emb.sin() + if rope_dims is not None and position_ids.shape[0] > 1: + mrope_section = rope_dims * 2 + cos = torch.cat([m[i % 3] for i, m in enumerate(cos.split(mrope_section, dim=-1))], dim=-1).unsqueeze(0) + sin = torch.cat([m[i % 3] for i, m in enumerate(sin.split(mrope_section, dim=-1))], dim=-1).unsqueeze(0) + else: + cos = cos.unsqueeze(1) + sin = sin.unsqueeze(1) + out.append((cos, sin)) + + if len(out) == 1: + return out[0] + + return out def apply_rope(xq, xk, freqs_cis): @@ -152,6 +203,14 @@ class Attention(nn.Module): self.v_proj = ops.Linear(config.hidden_size, self.num_kv_heads * self.head_dim, bias=config.qkv_bias, device=device, dtype=dtype) self.o_proj = ops.Linear(self.inner_size, config.hidden_size, bias=False, device=device, dtype=dtype) + self.q_norm = None + self.k_norm = None + + if config.q_norm == "gemma3": + self.q_norm = RMSNorm(self.head_dim, eps=config.rms_norm_eps, add=config.rms_norm_add, device=device, dtype=dtype) + if config.k_norm == "gemma3": + self.k_norm = RMSNorm(self.head_dim, eps=config.rms_norm_eps, add=config.rms_norm_add, device=device, dtype=dtype) + def forward( self, hidden_states: torch.Tensor, @@ -168,6 +227,11 @@ class Attention(nn.Module): xk = xk.view(batch_size, seq_length, self.num_kv_heads, self.head_dim).transpose(1, 2) xv = xv.view(batch_size, seq_length, self.num_kv_heads, self.head_dim).transpose(1, 2) + if self.q_norm is not None: + xq = self.q_norm(xq) + if self.k_norm is not None: + xk = self.k_norm(xk) + xq, xk = apply_rope(xq, xk, freqs_cis=freqs_cis) xk = xk.repeat_interleave(self.num_heads // self.num_kv_heads, dim=1) @@ -192,7 +256,7 @@ class MLP(nn.Module): return self.down_proj(self.activation(self.gate_proj(x)) * self.up_proj(x)) class TransformerBlock(nn.Module): - def __init__(self, config: Llama2Config, device=None, dtype=None, ops: Any = None): + def __init__(self, config: Llama2Config, index, device=None, dtype=None, ops: Any = None): super().__init__() self.self_attn = Attention(config, device=device, dtype=dtype, ops=ops) self.mlp = MLP(config, device=device, dtype=dtype, ops=ops) @@ -226,7 +290,7 @@ class TransformerBlock(nn.Module): return x class TransformerBlockGemma2(nn.Module): - def __init__(self, config: Llama2Config, device=None, dtype=None, ops: Any = None): + def __init__(self, config: Llama2Config, index, device=None, dtype=None, ops: Any = None): super().__init__() self.self_attn = Attention(config, device=device, dtype=dtype, ops=ops) self.mlp = MLP(config, device=device, dtype=dtype, ops=ops) @@ -235,6 +299,13 @@ class TransformerBlockGemma2(nn.Module): self.pre_feedforward_layernorm = RMSNorm(config.hidden_size, eps=config.rms_norm_eps, add=config.rms_norm_add, device=device, dtype=dtype) self.post_feedforward_layernorm = RMSNorm(config.hidden_size, eps=config.rms_norm_eps, add=config.rms_norm_add, device=device, dtype=dtype) + if config.sliding_attention is not None: # TODO: implement. (Not that necessary since models are trained on less than 1024 tokens) + self.sliding_attention = config.sliding_attention[index % len(config.sliding_attention)] + else: + self.sliding_attention = False + + self.transformer_type = config.transformer_type + def forward( self, x: torch.Tensor, @@ -242,6 +313,14 @@ class TransformerBlockGemma2(nn.Module): freqs_cis: Optional[torch.Tensor] = None, optimized_attention=None, ): + if self.transformer_type == 'gemma3': + if self.sliding_attention: + if x.shape[1] > self.sliding_attention: + logging.warning("Warning: sliding attention not implemented, results may be incorrect") + freqs_cis = freqs_cis[1] + else: + freqs_cis = freqs_cis[0] + # Self Attention residual = x x = self.input_layernorm(x) @@ -276,7 +355,7 @@ class Llama2_(nn.Module): device=device, dtype=dtype ) - if self.config.transformer_type == "gemma2": + if self.config.transformer_type == "gemma2" or self.config.transformer_type == "gemma3": transformer = TransformerBlockGemma2 self.normalize_in = True else: @@ -284,8 +363,8 @@ class Llama2_(nn.Module): self.normalize_in = False self.layers = nn.ModuleList([ - transformer(config, device=device, dtype=dtype, ops=ops) - for _ in range(config.num_hidden_layers) + transformer(config, index=i, device=device, dtype=dtype, ops=ops) + for i in range(config.num_hidden_layers) ]) self.norm = RMSNorm(config.hidden_size, eps=config.rms_norm_eps, add=config.rms_norm_add, device=device, dtype=dtype) # self.lm_head = ops.Linear(config.hidden_size, config.vocab_size, bias=False, device=device, dtype=dtype) @@ -305,6 +384,7 @@ class Llama2_(nn.Module): freqs_cis = precompute_freqs_cis(self.config.head_dim, position_ids, self.config.rope_theta, + self.config.rope_scale, self.config.rope_dims, device=x.device) @@ -433,3 +513,12 @@ class Gemma2_2B(BaseLlama, torch.nn.Module): self.model = Llama2_(config, device=device, dtype=dtype, ops=operations) self.dtype = dtype + +class Gemma3_4B(BaseLlama, torch.nn.Module): + def __init__(self, config_dict, dtype, device, operations): + super().__init__() + config = Gemma3_4B_Config(**config_dict) + self.num_layers = config.num_hidden_layers + + self.model = Llama2_(config, device=device, dtype=dtype, ops=operations) + self.dtype = dtype diff --git a/comfy/text_encoders/lumina2.py b/comfy/text_encoders/lumina2.py index 674461b75..fd986e2c1 100644 --- a/comfy/text_encoders/lumina2.py +++ b/comfy/text_encoders/lumina2.py @@ -11,23 +11,41 @@ class Gemma2BTokenizer(sd1_clip.SDTokenizer): def state_dict(self): return {"spiece_model": self.tokenizer.serialize_model()} +class Gemma3_4BTokenizer(sd1_clip.SDTokenizer): + def __init__(self, embedding_directory=None, tokenizer_data={}): + tokenizer = tokenizer_data.get("spiece_model", None) + super().__init__(tokenizer, pad_with_end=False, embedding_size=2560, embedding_key='gemma3_4b', tokenizer_class=SPieceTokenizer, has_end_token=False, pad_to_max_length=False, max_length=99999999, min_length=1, tokenizer_args={"add_bos": True, "add_eos": False}, tokenizer_data=tokenizer_data) + + def state_dict(self): + return {"spiece_model": self.tokenizer.serialize_model()} class LuminaTokenizer(sd1_clip.SD1Tokenizer): def __init__(self, embedding_directory=None, tokenizer_data={}): super().__init__(embedding_directory=embedding_directory, tokenizer_data=tokenizer_data, name="gemma2_2b", tokenizer=Gemma2BTokenizer) +class NTokenizer(sd1_clip.SD1Tokenizer): + def __init__(self, embedding_directory=None, tokenizer_data={}): + super().__init__(embedding_directory=embedding_directory, tokenizer_data=tokenizer_data, name="gemma3_4b", tokenizer=Gemma3_4BTokenizer) class Gemma2_2BModel(sd1_clip.SDClipModel): def __init__(self, device="cpu", layer="hidden", layer_idx=-2, dtype=None, attention_mask=True, model_options={}): super().__init__(device=device, layer=layer, layer_idx=layer_idx, textmodel_json_config={}, dtype=dtype, special_tokens={"start": 2, "pad": 0}, layer_norm_hidden_state=False, model_class=comfy.text_encoders.llama.Gemma2_2B, enable_attention_masks=attention_mask, return_attention_masks=attention_mask, model_options=model_options) +class Gemma3_4BModel(sd1_clip.SDClipModel): + def __init__(self, device="cpu", layer="hidden", layer_idx=-2, dtype=None, attention_mask=True, model_options={}): + super().__init__(device=device, layer=layer, layer_idx=layer_idx, textmodel_json_config={}, dtype=dtype, special_tokens={"start": 2, "pad": 0}, layer_norm_hidden_state=False, model_class=comfy.text_encoders.llama.Gemma3_4B, enable_attention_masks=attention_mask, return_attention_masks=attention_mask, model_options=model_options) class LuminaModel(sd1_clip.SD1ClipModel): - def __init__(self, device="cpu", dtype=None, model_options={}): - super().__init__(device=device, dtype=dtype, name="gemma2_2b", clip_model=Gemma2_2BModel, model_options=model_options) + def __init__(self, device="cpu", dtype=None, model_options={}, name="gemma2_2b", clip_model=Gemma2_2BModel): + super().__init__(device=device, dtype=dtype, name=name, clip_model=clip_model, model_options=model_options) -def te(dtype_llama=None, llama_scaled_fp8=None): +def te(dtype_llama=None, llama_scaled_fp8=None, model_type="gemma2_2b"): + if model_type == "gemma2_2b": + model = Gemma2_2BModel + elif model_type == "gemma3_4b": + model = Gemma3_4BModel + class LuminaTEModel_(LuminaModel): def __init__(self, device="cpu", dtype=None, model_options={}): if llama_scaled_fp8 is not None and "scaled_fp8" not in model_options: @@ -35,5 +53,5 @@ def te(dtype_llama=None, llama_scaled_fp8=None): model_options["scaled_fp8"] = llama_scaled_fp8 if dtype_llama is not None: dtype = dtype_llama - super().__init__(device=device, dtype=dtype, model_options=model_options) + super().__init__(device=device, dtype=dtype, name=model_type, model_options=model_options, clip_model=model) return LuminaTEModel_ From fc34c3d1125e970699dcb311323839ed6dda4985 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Tue, 7 Oct 2025 23:15:32 +0300 Subject: [PATCH 108/114] fix(ReCraft-API-node): allow custom multipart parser to return FormData (#10244) --- comfy_api_nodes/apis/client.py | 17 ++++++++++------- comfy_api_nodes/nodes_recraft.py | 28 ++++++++++++++++++++++------ 2 files changed, 32 insertions(+), 13 deletions(-) diff --git a/comfy_api_nodes/apis/client.py b/comfy_api_nodes/apis/client.py index 79de3c262..36560c9e3 100644 --- a/comfy_api_nodes/apis/client.py +++ b/comfy_api_nodes/apis/client.py @@ -220,13 +220,16 @@ class ApiClient: if multipart_parser and data: data = multipart_parser(data) - form = aiohttp.FormData(default_to_multipart=True) - if data: # regular text fields - for k, v in data.items(): - if v is None: - continue # aiohttp fails to serialize "None" values - # aiohttp expects strings or bytes; convert enums etc. - form.add_field(k, str(v) if not isinstance(v, (bytes, bytearray)) else v) + if isinstance(data, aiohttp.FormData): + form = data # If the parser already returned a FormData, pass it through + else: + form = aiohttp.FormData(default_to_multipart=True) + if data: # regular text fields + for k, v in data.items(): + if v is None: + continue # aiohttp fails to serialize "None" values + # aiohttp expects strings or bytes; convert enums etc. + form.add_field(k, str(v) if not isinstance(v, (bytes, bytearray)) else v) if files: file_iter = files if isinstance(files, list) else files.items() diff --git a/comfy_api_nodes/nodes_recraft.py b/comfy_api_nodes/nodes_recraft.py index 0bbb551b8..8beed5675 100644 --- a/comfy_api_nodes/nodes_recraft.py +++ b/comfy_api_nodes/nodes_recraft.py @@ -35,6 +35,7 @@ from server import PromptServer import torch from io import BytesIO from PIL import UnidentifiedImageError +import aiohttp async def handle_recraft_file_request( @@ -82,10 +83,16 @@ async def handle_recraft_file_request( return all_bytesio -def recraft_multipart_parser(data, parent_key=None, formatter: callable=None, converted_to_check: list[list]=None, is_list=False) -> dict: +def recraft_multipart_parser( + data, + parent_key=None, + formatter: callable = None, + converted_to_check: list[list] = None, + is_list: bool = False, + return_mode: str = "formdata" # "dict" | "formdata" +) -> dict | aiohttp.FormData: """ - Formats data such that multipart/form-data will work with requests library - when both files and data are present. + Formats data such that multipart/form-data will work with aiohttp library when both files and data are present. The OpenAI client that Recraft uses has a bizarre way of serializing lists: @@ -103,19 +110,19 @@ def recraft_multipart_parser(data, parent_key=None, formatter: callable=None, co # Modification of a function that handled a different type of multipart parsing, big ups: # https://gist.github.com/kazqvaizer/4cebebe5db654a414132809f9f88067b - def handle_converted_lists(data, parent_key, lists_to_check=tuple[list]): + def handle_converted_lists(item, parent_key, lists_to_check=tuple[list]): # if list already exists exists, just extend list with data for check_list in lists_to_check: for conv_tuple in check_list: if conv_tuple[0] == parent_key and isinstance(conv_tuple[1], list): - conv_tuple[1].append(formatter(data)) + conv_tuple[1].append(formatter(item)) return True return False if converted_to_check is None: converted_to_check = [] - + effective_mode = return_mode if parent_key is None else "dict" if formatter is None: formatter = lambda v: v # Multipart representation of value @@ -145,6 +152,15 @@ def recraft_multipart_parser(data, parent_key=None, formatter: callable=None, co else: converted.append((current_key, formatter(value))) + if effective_mode == "formdata": + fd = aiohttp.FormData() + for k, v in dict(converted).items(): + if isinstance(v, list): + for item in v: + fd.add_field(k, str(item)) + else: + fd.add_field(k, str(v)) + return fd return dict(converted) From 9e984c48bc6a1d1c82231c46542995dbf5a265d7 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 8 Oct 2025 00:11:37 +0300 Subject: [PATCH 109/114] feat(api-nodes): add Sora2 API node (#10249) --- comfy_api_nodes/apinode_utils.py | 23 +++- comfy_api_nodes/nodes_sora.py | 175 +++++++++++++++++++++++++++++++ nodes.py | 1 + 3 files changed, 194 insertions(+), 5 deletions(-) create mode 100644 comfy_api_nodes/nodes_sora.py diff --git a/comfy_api_nodes/apinode_utils.py b/comfy_api_nodes/apinode_utils.py index 5ac3b92aa..571c74286 100644 --- a/comfy_api_nodes/apinode_utils.py +++ b/comfy_api_nodes/apinode_utils.py @@ -18,7 +18,7 @@ from comfy_api_nodes.apis.client import ( UploadResponse, ) from server import PromptServer - +from comfy.cli_args import args import numpy as np from PIL import Image @@ -30,7 +30,9 @@ from io import BytesIO import av -async def download_url_to_video_output(video_url: str, timeout: int = None) -> VideoFromFile: +async def download_url_to_video_output( + video_url: str, timeout: int = None, auth_kwargs: Optional[dict[str, str]] = None +) -> VideoFromFile: """Downloads a video from a URL and returns a `VIDEO` output. Args: @@ -39,7 +41,7 @@ async def download_url_to_video_output(video_url: str, timeout: int = None) -> V Returns: A Comfy node `VIDEO` output. """ - video_io = await download_url_to_bytesio(video_url, timeout) + video_io = await download_url_to_bytesio(video_url, timeout, auth_kwargs=auth_kwargs) if video_io is None: error_msg = f"Failed to download video from {video_url}" logging.error(error_msg) @@ -164,7 +166,9 @@ def mimetype_to_extension(mime_type: str) -> str: return mime_type.split("/")[-1].lower() -async def download_url_to_bytesio(url: str, timeout: int = None) -> BytesIO: +async def download_url_to_bytesio( + url: str, timeout: int = None, auth_kwargs: Optional[dict[str, str]] = None +) -> BytesIO: """Downloads content from a URL using requests and returns it as BytesIO. Args: @@ -174,9 +178,18 @@ async def download_url_to_bytesio(url: str, timeout: int = None) -> BytesIO: Returns: BytesIO object containing the downloaded content. """ + headers = {} + if url.startswith("/proxy/"): + url = str(args.comfy_api_base).rstrip("/") + url + auth_token = auth_kwargs.get("auth_token") + comfy_api_key = auth_kwargs.get("comfy_api_key") + if auth_token: + headers["Authorization"] = f"Bearer {auth_token}" + elif comfy_api_key: + headers["X-API-KEY"] = comfy_api_key timeout_cfg = aiohttp.ClientTimeout(total=timeout) if timeout else None async with aiohttp.ClientSession(timeout=timeout_cfg) as session: - async with session.get(url) as resp: + async with session.get(url, headers=headers) as resp: resp.raise_for_status() # Raises HTTPError for bad responses (4XX or 5XX) return BytesIO(await resp.read()) diff --git a/comfy_api_nodes/nodes_sora.py b/comfy_api_nodes/nodes_sora.py new file mode 100644 index 000000000..2d532d637 --- /dev/null +++ b/comfy_api_nodes/nodes_sora.py @@ -0,0 +1,175 @@ +from typing import Optional +from typing_extensions import override + +import torch +from pydantic import BaseModel, Field +from comfy_api.latest import ComfyExtension, io as comfy_io +from comfy_api_nodes.apis.client import ( + ApiEndpoint, + HttpMethod, + SynchronousOperation, + PollingOperation, + EmptyRequest, +) +from comfy_api_nodes.util.validation_utils import get_number_of_images + +from comfy_api_nodes.apinode_utils import ( + download_url_to_video_output, + tensor_to_bytesio, +) + +class Sora2GenerationRequest(BaseModel): + prompt: str = Field(...) + model: str = Field(...) + seconds: str = Field(...) + size: str = Field(...) + + +class Sora2GenerationResponse(BaseModel): + id: str = Field(...) + error: Optional[dict] = Field(None) + status: Optional[str] = Field(None) + + +class OpenAIVideoSora2(comfy_io.ComfyNode): + @classmethod + def define_schema(cls): + return comfy_io.Schema( + node_id="OpenAIVideoSora2", + display_name="OpenAI Sora - Video", + category="api node/video/Sora", + description="OpenAI video and audio generation.", + inputs=[ + comfy_io.Combo.Input( + "model", + options=["sora-2", "sora-2-pro"], + default="sora-2", + ), + comfy_io.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Guiding text; may be empty if an input image is present.", + ), + comfy_io.Combo.Input( + "size", + options=[ + "720x1280", + "1280x720", + "1024x1792", + "1792x1024", + ], + default="1280x720", + ), + comfy_io.Combo.Input( + "duration", + options=[4, 8, 12], + default=8, + ), + comfy_io.Image.Input( + "image", + optional=True, + ), + comfy_io.Int.Input( + "seed", + default=0, + min=0, + max=2147483647, + step=1, + display_mode=comfy_io.NumberDisplay.number, + control_after_generate=True, + optional=True, + tooltip="Seed to determine if node should re-run; " + "actual results are nondeterministic regardless of seed.", + ), + ], + outputs=[ + comfy_io.Video.Output(), + ], + hidden=[ + comfy_io.Hidden.auth_token_comfy_org, + comfy_io.Hidden.api_key_comfy_org, + comfy_io.Hidden.unique_id, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, + model: str, + prompt: str, + size: str = "1280x720", + duration: int = 8, + seed: int = 0, + image: Optional[torch.Tensor] = None, + ): + if model == "sora-2" and size not in ("720x1280", "1280x720"): + raise ValueError("Invalid size for sora-2 model, only 720x1280 and 1280x720 are supported.") + files_input = None + if image is not None: + if get_number_of_images(image) != 1: + raise ValueError("Currently only one input image is supported.") + files_input = {"input_reference": ("image.png", tensor_to_bytesio(image), "image/png")} + auth = { + "auth_token": cls.hidden.auth_token_comfy_org, + "comfy_api_key": cls.hidden.api_key_comfy_org, + } + payload = Sora2GenerationRequest( + model=model, + prompt=prompt, + seconds=str(duration), + size=size, + ) + initial_operation = SynchronousOperation( + endpoint=ApiEndpoint( + path="/proxy/openai/v1/videos", + method=HttpMethod.POST, + request_model=Sora2GenerationRequest, + response_model=Sora2GenerationResponse + ), + request=payload, + files=files_input, + auth_kwargs=auth, + content_type="multipart/form-data", + ) + initial_response = await initial_operation.execute() + if initial_response.error: + raise Exception(initial_response.error.message) + + model_time_multiplier = 1 if model == "sora-2" else 2 + poll_operation = PollingOperation( + poll_endpoint=ApiEndpoint( + path=f"/proxy/openai/v1/videos/{initial_response.id}", + method=HttpMethod.GET, + request_model=EmptyRequest, + response_model=Sora2GenerationResponse + ), + completed_statuses=["completed"], + failed_statuses=["failed"], + status_extractor=lambda x: x.status, + auth_kwargs=auth, + poll_interval=8.0, + max_poll_attempts=160, + node_id=cls.hidden.unique_id, + estimated_duration=45 * (duration / 4) * model_time_multiplier, + ) + await poll_operation.execute() + return comfy_io.NodeOutput( + await download_url_to_video_output( + f"/proxy/openai/v1/videos/{initial_response.id}/content", + auth_kwargs=auth, + ) + ) + + +class OpenAISoraExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[comfy_io.ComfyNode]]: + return [ + OpenAIVideoSora2, + ] + + +async def comfy_entrypoint() -> OpenAISoraExtension: + return OpenAISoraExtension() diff --git a/nodes.py b/nodes.py index 88d712993..2a2a5f2ad 100644 --- a/nodes.py +++ b/nodes.py @@ -2357,6 +2357,7 @@ async def init_builtin_api_nodes(): "nodes_stability.py", "nodes_pika.py", "nodes_runway.py", + "nodes_sora.py", "nodes_tripo.py", "nodes_moonvalley.py", "nodes_rodin.py", From 8a15568f10c0622a7281c32fadffc51511e53c10 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Tue, 7 Oct 2025 16:55:23 -0700 Subject: [PATCH 110/114] Temp fix for LTXV custom nodes. (#10251) --- comfy_extras/nodes_lt.py | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/comfy_extras/nodes_lt.py b/comfy_extras/nodes_lt.py index b51d15804..50da5f4eb 100644 --- a/comfy_extras/nodes_lt.py +++ b/comfy_extras/nodes_lt.py @@ -34,6 +34,7 @@ class EmptyLTXVLatentVideo(io.ComfyNode): latent = torch.zeros([batch_size, 128, ((length - 1) // 8) + 1, height // 32, width // 32], device=comfy.model_management.intermediate_device()) return io.NodeOutput({"samples": latent}) + generate = execute # TODO: remove class LTXVImgToVideo(io.ComfyNode): @classmethod @@ -77,6 +78,8 @@ class LTXVImgToVideo(io.ComfyNode): return io.NodeOutput(positive, negative, {"samples": latent, "noise_mask": conditioning_latent_frames_mask}) + generate = execute # TODO: remove + def conditioning_get_any_value(conditioning, key, default=None): for t in conditioning: @@ -264,6 +267,8 @@ class LTXVAddGuide(io.ComfyNode): return io.NodeOutput(positive, negative, {"samples": latent_image, "noise_mask": noise_mask}) + generate = execute # TODO: remove + class LTXVCropGuides(io.ComfyNode): @classmethod @@ -300,6 +305,8 @@ class LTXVCropGuides(io.ComfyNode): return io.NodeOutput(positive, negative, {"samples": latent_image, "noise_mask": noise_mask}) + crop = execute # TODO: remove + class LTXVConditioning(io.ComfyNode): @classmethod @@ -498,6 +505,7 @@ class LTXVPreprocess(io.ComfyNode): output_images.append(preprocess(image[i], img_compression)) return io.NodeOutput(torch.stack(output_images)) + preprocess = execute # TODO: remove class LtxvExtension(ComfyExtension): @override From 19f595b788bd227004a5f7232f3b5895b46411ea Mon Sep 17 00:00:00 2001 From: filtered <176114999+webfiltered@users.noreply.github.com> Date: Wed, 8 Oct 2025 11:54:00 +1100 Subject: [PATCH 111/114] Bump frontend to 1.27.10 (#10252) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index db0486960..85b3bb63b 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,4 +1,4 @@ -comfyui-frontend-package==1.27.7 +comfyui-frontend-package==1.27.10 comfyui-workflow-templates==0.1.93 comfyui-embedded-docs==0.2.6 torch From 51697d50dc94005b1c279eb0cf45207697946020 Mon Sep 17 00:00:00 2001 From: ComfyUI Wiki Date: Wed, 8 Oct 2025 10:48:51 +0800 Subject: [PATCH 112/114] update template to 0.1.94 (#10253) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 85b3bb63b..d4594df39 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,5 +1,5 @@ comfyui-frontend-package==1.27.10 -comfyui-workflow-templates==0.1.93 +comfyui-workflow-templates==0.1.94 comfyui-embedded-docs==0.2.6 torch torchsde From 637221995f7424a561bd825de3e61ea117dfe1e3 Mon Sep 17 00:00:00 2001 From: comfyanonymous Date: Wed, 8 Oct 2025 00:53:43 -0400 Subject: [PATCH 113/114] ComfyUI version 0.3.64 --- comfyui_version.py | 2 +- pyproject.toml | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/comfyui_version.py b/comfyui_version.py index c3257d4bf..da5cde02d 100644 --- a/comfyui_version.py +++ b/comfyui_version.py @@ -1,3 +1,3 @@ # This file is automatically generated by the build process when version is # updated in pyproject.toml. -__version__ = "0.3.63" +__version__ = "0.3.64" diff --git a/pyproject.toml b/pyproject.toml index abd1a5f5c..6ea839336 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -1,6 +1,6 @@ [project] name = "ComfyUI" -version = "0.3.63" +version = "0.3.64" readme = "README.md" license = { file = "LICENSE" } requires-python = ">=3.9" From 3e0eb8d33f9a65f2a01430f1b4a1535348af881c Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Wed, 8 Oct 2025 10:14:04 +0300 Subject: [PATCH 114/114] feat(V3-io): allow Enum classes for Combo options (#10237) --- comfy_api/latest/_io.py | 24 ++++++++++++++++----- comfy_api_nodes/nodes_bytedance.py | 16 +++++++------- comfy_api_nodes/nodes_kling.py | 20 +++++++++--------- comfy_api_nodes/nodes_luma.py | 18 ++++++++-------- comfy_api_nodes/nodes_pika.py | 6 +++--- comfy_api_nodes/nodes_pixverse.py | 22 +++++++++---------- comfy_api_nodes/nodes_runway.py | 12 +++++------ comfy_api_nodes/nodes_stability.py | 10 ++++----- comfy_api_nodes/nodes_vidu.py | 34 +++++++++++++++--------------- 9 files changed, 88 insertions(+), 74 deletions(-) diff --git a/comfy_api/latest/_io.py b/comfy_api/latest/_io.py index 2d95cffd6..661309f19 100644 --- a/comfy_api/latest/_io.py +++ b/comfy_api/latest/_io.py @@ -336,11 +336,25 @@ class Combo(ComfyTypeIO): class Input(WidgetInput): """Combo input (dropdown).""" Type = str - def __init__(self, id: str, options: list[str]=None, display_name: str=None, optional=False, tooltip: str=None, lazy: bool=None, - default: str=None, control_after_generate: bool=None, - upload: UploadType=None, image_folder: FolderType=None, - remote: RemoteOptions=None, - socketless: bool=None): + def __init__( + self, + id: str, + options: list[str] | list[int] | type[Enum] = None, + display_name: str=None, + optional=False, + tooltip: str=None, + lazy: bool=None, + default: str | int | Enum = None, + control_after_generate: bool=None, + upload: UploadType=None, + image_folder: FolderType=None, + remote: RemoteOptions=None, + socketless: bool=None, + ): + if isinstance(options, type) and issubclass(options, Enum): + options = [v.value for v in options] + if isinstance(default, Enum): + default = default.value super().__init__(id, display_name, optional, tooltip, lazy, default, socketless) self.multiselect = False self.options = options diff --git a/comfy_api_nodes/nodes_bytedance.py b/comfy_api_nodes/nodes_bytedance.py index 654d6a362..fcb01820c 100644 --- a/comfy_api_nodes/nodes_bytedance.py +++ b/comfy_api_nodes/nodes_bytedance.py @@ -249,8 +249,8 @@ class ByteDanceImageNode(comfy_io.ComfyNode): inputs=[ comfy_io.Combo.Input( "model", - options=[model.value for model in Text2ImageModelName], - default=Text2ImageModelName.seedream_3.value, + options=Text2ImageModelName, + default=Text2ImageModelName.seedream_3, tooltip="Model name", ), comfy_io.String.Input( @@ -382,8 +382,8 @@ class ByteDanceImageEditNode(comfy_io.ComfyNode): inputs=[ comfy_io.Combo.Input( "model", - options=[model.value for model in Image2ImageModelName], - default=Image2ImageModelName.seededit_3.value, + options=Image2ImageModelName, + default=Image2ImageModelName.seededit_3, tooltip="Model name", ), comfy_io.Image.Input( @@ -676,8 +676,8 @@ class ByteDanceTextToVideoNode(comfy_io.ComfyNode): inputs=[ comfy_io.Combo.Input( "model", - options=[model.value for model in Text2VideoModelName], - default=Text2VideoModelName.seedance_1_pro.value, + options=Text2VideoModelName, + default=Text2VideoModelName.seedance_1_pro, tooltip="Model name", ), comfy_io.String.Input( @@ -793,8 +793,8 @@ class ByteDanceImageToVideoNode(comfy_io.ComfyNode): inputs=[ comfy_io.Combo.Input( "model", - options=[model.value for model in Image2VideoModelName], - default=Image2VideoModelName.seedance_1_pro.value, + options=Image2VideoModelName, + default=Image2VideoModelName.seedance_1_pro, tooltip="Model name", ), comfy_io.String.Input( diff --git a/comfy_api_nodes/nodes_kling.py b/comfy_api_nodes/nodes_kling.py index 457b43451..fe5b8562d 100644 --- a/comfy_api_nodes/nodes_kling.py +++ b/comfy_api_nodes/nodes_kling.py @@ -647,7 +647,7 @@ class KlingCameraControls(comfy_io.ComfyNode): category="api node/video/Kling", description="Allows specifying configuration options for Kling Camera Controls and motion control effects.", inputs=[ - comfy_io.Combo.Input("camera_control_type", options=[i.value for i in KlingCameraControlType]), + comfy_io.Combo.Input("camera_control_type", options=KlingCameraControlType), comfy_io.Float.Input( "horizontal_movement", default=0.0, @@ -772,7 +772,7 @@ class KlingTextToVideoNode(comfy_io.ComfyNode): comfy_io.Float.Input("cfg_scale", default=1.0, min=0.0, max=1.0), comfy_io.Combo.Input( "aspect_ratio", - options=[i.value for i in KlingVideoGenAspectRatio], + options=KlingVideoGenAspectRatio, default="16:9", ), comfy_io.Combo.Input( @@ -840,7 +840,7 @@ class KlingCameraControlT2VNode(comfy_io.ComfyNode): comfy_io.Float.Input("cfg_scale", default=0.75, min=0.0, max=1.0), comfy_io.Combo.Input( "aspect_ratio", - options=[i.value for i in KlingVideoGenAspectRatio], + options=KlingVideoGenAspectRatio, default="16:9", ), comfy_io.Custom("CAMERA_CONTROL").Input( @@ -903,17 +903,17 @@ class KlingImage2VideoNode(comfy_io.ComfyNode): comfy_io.String.Input("negative_prompt", multiline=True, tooltip="Negative text prompt"), comfy_io.Combo.Input( "model_name", - options=[i.value for i in KlingVideoGenModelName], + options=KlingVideoGenModelName, default="kling-v2-master", ), comfy_io.Float.Input("cfg_scale", default=0.8, min=0.0, max=1.0), - comfy_io.Combo.Input("mode", options=[i.value for i in KlingVideoGenMode], default="std"), + comfy_io.Combo.Input("mode", options=KlingVideoGenMode, default=KlingVideoGenMode.std), comfy_io.Combo.Input( "aspect_ratio", - options=[i.value for i in KlingVideoGenAspectRatio], - default="16:9", + options=KlingVideoGenAspectRatio, + default=KlingVideoGenAspectRatio.field_16_9, ), - comfy_io.Combo.Input("duration", options=[i.value for i in KlingVideoGenDuration], default="5"), + comfy_io.Combo.Input("duration", options=KlingVideoGenDuration, default=KlingVideoGenDuration.field_5), ], outputs=[ comfy_io.Video.Output(), @@ -984,8 +984,8 @@ class KlingCameraControlI2VNode(comfy_io.ComfyNode): comfy_io.Float.Input("cfg_scale", default=0.75, min=0.0, max=1.0), comfy_io.Combo.Input( "aspect_ratio", - options=[i.value for i in KlingVideoGenAspectRatio], - default="16:9", + options=KlingVideoGenAspectRatio, + default=KlingVideoGenAspectRatio.field_16_9, ), comfy_io.Custom("CAMERA_CONTROL").Input( "camera_control", diff --git a/comfy_api_nodes/nodes_luma.py b/comfy_api_nodes/nodes_luma.py index 9cd02ffd2..9cab2ca82 100644 --- a/comfy_api_nodes/nodes_luma.py +++ b/comfy_api_nodes/nodes_luma.py @@ -181,11 +181,11 @@ class LumaImageGenerationNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "model", - options=[model.value for model in LumaImageModel], + options=LumaImageModel, ), comfy_io.Combo.Input( "aspect_ratio", - options=[ratio.value for ratio in LumaAspectRatio], + options=LumaAspectRatio, default=LumaAspectRatio.ratio_16_9, ), comfy_io.Int.Input( @@ -366,7 +366,7 @@ class LumaImageModifyNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "model", - options=[model.value for model in LumaImageModel], + options=LumaImageModel, ), comfy_io.Int.Input( "seed", @@ -466,21 +466,21 @@ class LumaTextToVideoGenerationNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "model", - options=[model.value for model in LumaVideoModel], + options=LumaVideoModel, ), comfy_io.Combo.Input( "aspect_ratio", - options=[ratio.value for ratio in LumaAspectRatio], + options=LumaAspectRatio, default=LumaAspectRatio.ratio_16_9, ), comfy_io.Combo.Input( "resolution", - options=[resolution.value for resolution in LumaVideoOutputResolution], + options=LumaVideoOutputResolution, default=LumaVideoOutputResolution.res_540p, ), comfy_io.Combo.Input( "duration", - options=[dur.value for dur in LumaVideoModelOutputDuration], + options=LumaVideoModelOutputDuration, ), comfy_io.Boolean.Input( "loop", @@ -595,7 +595,7 @@ class LumaImageToVideoGenerationNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "model", - options=[model.value for model in LumaVideoModel], + options=LumaVideoModel, ), # comfy_io.Combo.Input( # "aspect_ratio", @@ -604,7 +604,7 @@ class LumaImageToVideoGenerationNode(comfy_io.ComfyNode): # ), comfy_io.Combo.Input( "resolution", - options=[resolution.value for resolution in LumaVideoOutputResolution], + options=LumaVideoOutputResolution, default=LumaVideoOutputResolution.res_540p, ), comfy_io.Combo.Input( diff --git a/comfy_api_nodes/nodes_pika.py b/comfy_api_nodes/nodes_pika.py index 0a9f04cc2..35d6baf1c 100644 --- a/comfy_api_nodes/nodes_pika.py +++ b/comfy_api_nodes/nodes_pika.py @@ -174,10 +174,10 @@ def get_base_inputs_types() -> list[comfy_io.Input]: comfy_io.String.Input("negative_prompt", multiline=True), comfy_io.Int.Input("seed", min=0, max=0xFFFFFFFF, control_after_generate=True), comfy_io.Combo.Input( - "resolution", options=[resolution.value for resolution in PikaResolutionEnum], default="1080p" + "resolution", options=PikaResolutionEnum, default=PikaResolutionEnum.field_1080p ), comfy_io.Combo.Input( - "duration", options=[duration.value for duration in PikaDurationEnum], default=5 + "duration", options=PikaDurationEnum, default=PikaDurationEnum.integer_5 ), ] @@ -616,7 +616,7 @@ class PikaffectsNode(comfy_io.ComfyNode): inputs=[ comfy_io.Image.Input("image", tooltip="The reference image to apply the Pikaffect to."), comfy_io.Combo.Input( - "pikaffect", options=[pikaffect.value for pikaffect in Pikaffect], default="Cake-ify" + "pikaffect", options=Pikaffect, default="Cake-ify" ), comfy_io.String.Input("prompt_text", multiline=True), comfy_io.String.Input("negative_prompt", multiline=True), diff --git a/comfy_api_nodes/nodes_pixverse.py b/comfy_api_nodes/nodes_pixverse.py index 2c91bbc65..a97610f06 100644 --- a/comfy_api_nodes/nodes_pixverse.py +++ b/comfy_api_nodes/nodes_pixverse.py @@ -85,7 +85,7 @@ class PixverseTemplateNode(comfy_io.ComfyNode): display_name="PixVerse Template", category="api node/video/PixVerse", inputs=[ - comfy_io.Combo.Input("template", options=[list(pixverse_templates.keys())]), + comfy_io.Combo.Input("template", options=list(pixverse_templates.keys())), ], outputs=[comfy_io.Custom(PixverseIO.TEMPLATE).Output(display_name="pixverse_template")], ) @@ -120,20 +120,20 @@ class PixverseTextToVideoNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "aspect_ratio", - options=[ratio.value for ratio in PixverseAspectRatio], + options=PixverseAspectRatio, ), comfy_io.Combo.Input( "quality", - options=[resolution.value for resolution in PixverseQuality], + options=PixverseQuality, default=PixverseQuality.res_540p, ), comfy_io.Combo.Input( "duration_seconds", - options=[dur.value for dur in PixverseDuration], + options=PixverseDuration, ), comfy_io.Combo.Input( "motion_mode", - options=[mode.value for mode in PixverseMotionMode], + options=PixverseMotionMode, ), comfy_io.Int.Input( "seed", @@ -262,16 +262,16 @@ class PixverseImageToVideoNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "quality", - options=[resolution.value for resolution in PixverseQuality], + options=PixverseQuality, default=PixverseQuality.res_540p, ), comfy_io.Combo.Input( "duration_seconds", - options=[dur.value for dur in PixverseDuration], + options=PixverseDuration, ), comfy_io.Combo.Input( "motion_mode", - options=[mode.value for mode in PixverseMotionMode], + options=PixverseMotionMode, ), comfy_io.Int.Input( "seed", @@ -403,16 +403,16 @@ class PixverseTransitionVideoNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "quality", - options=[resolution.value for resolution in PixverseQuality], + options=PixverseQuality, default=PixverseQuality.res_540p, ), comfy_io.Combo.Input( "duration_seconds", - options=[dur.value for dur in PixverseDuration], + options=PixverseDuration, ), comfy_io.Combo.Input( "motion_mode", - options=[mode.value for mode in PixverseMotionMode], + options=PixverseMotionMode, ), comfy_io.Int.Input( "seed", diff --git a/comfy_api_nodes/nodes_runway.py b/comfy_api_nodes/nodes_runway.py index 27b2bf748..ea22692cb 100644 --- a/comfy_api_nodes/nodes_runway.py +++ b/comfy_api_nodes/nodes_runway.py @@ -200,11 +200,11 @@ class RunwayImageToVideoNodeGen3a(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "duration", - options=[model.value for model in Duration], + options=Duration, ), comfy_io.Combo.Input( "ratio", - options=[model.value for model in RunwayGen3aAspectRatio], + options=RunwayGen3aAspectRatio, ), comfy_io.Int.Input( "seed", @@ -300,11 +300,11 @@ class RunwayImageToVideoNodeGen4(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "duration", - options=[model.value for model in Duration], + options=Duration, ), comfy_io.Combo.Input( "ratio", - options=[model.value for model in RunwayGen4TurboAspectRatio], + options=RunwayGen4TurboAspectRatio, ), comfy_io.Int.Input( "seed", @@ -408,11 +408,11 @@ class RunwayFirstLastFrameNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "duration", - options=[model.value for model in Duration], + options=Duration, ), comfy_io.Combo.Input( "ratio", - options=[model.value for model in RunwayGen3aAspectRatio], + options=RunwayGen3aAspectRatio, ), comfy_io.Int.Input( "seed", diff --git a/comfy_api_nodes/nodes_stability.py b/comfy_api_nodes/nodes_stability.py index 5ba5ed986..bfb67fc9d 100644 --- a/comfy_api_nodes/nodes_stability.py +++ b/comfy_api_nodes/nodes_stability.py @@ -82,8 +82,8 @@ class StabilityStableImageUltraNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "aspect_ratio", - options=[x.value for x in StabilityAspectRatio], - default=StabilityAspectRatio.ratio_1_1.value, + options=StabilityAspectRatio, + default=StabilityAspectRatio.ratio_1_1, tooltip="Aspect ratio of generated image.", ), comfy_io.Combo.Input( @@ -217,12 +217,12 @@ class StabilityStableImageSD_3_5Node(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "model", - options=[x.value for x in Stability_SD3_5_Model], + options=Stability_SD3_5_Model, ), comfy_io.Combo.Input( "aspect_ratio", - options=[x.value for x in StabilityAspectRatio], - default=StabilityAspectRatio.ratio_1_1.value, + options=StabilityAspectRatio, + default=StabilityAspectRatio.ratio_1_1, tooltip="Aspect ratio of generated image.", ), comfy_io.Combo.Input( diff --git a/comfy_api_nodes/nodes_vidu.py b/comfy_api_nodes/nodes_vidu.py index 2f441948c..ac28b683c 100644 --- a/comfy_api_nodes/nodes_vidu.py +++ b/comfy_api_nodes/nodes_vidu.py @@ -173,8 +173,8 @@ class ViduTextToVideoNode(comfy_io.ComfyNode): inputs=[ comfy_io.Combo.Input( "model", - options=[model.value for model in VideoModelName], - default=VideoModelName.vidu_q1.value, + options=VideoModelName, + default=VideoModelName.vidu_q1, tooltip="Model name", ), comfy_io.String.Input( @@ -205,22 +205,22 @@ class ViduTextToVideoNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "aspect_ratio", - options=[model.value for model in AspectRatio], - default=AspectRatio.r_16_9.value, + options=AspectRatio, + default=AspectRatio.r_16_9, tooltip="The aspect ratio of the output video", optional=True, ), comfy_io.Combo.Input( "resolution", - options=[model.value for model in Resolution], - default=Resolution.r_1080p.value, + options=Resolution, + default=Resolution.r_1080p, tooltip="Supported values may vary by model & duration", optional=True, ), comfy_io.Combo.Input( "movement_amplitude", - options=[model.value for model in MovementAmplitude], - default=MovementAmplitude.auto.value, + options=MovementAmplitude, + default=MovementAmplitude.auto, tooltip="The movement amplitude of objects in the frame", optional=True, ), @@ -278,8 +278,8 @@ class ViduImageToVideoNode(comfy_io.ComfyNode): inputs=[ comfy_io.Combo.Input( "model", - options=[model.value for model in VideoModelName], - default=VideoModelName.vidu_q1.value, + options=VideoModelName, + default=VideoModelName.vidu_q1, tooltip="Model name", ), comfy_io.Image.Input( @@ -316,14 +316,14 @@ class ViduImageToVideoNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "resolution", - options=[model.value for model in Resolution], - default=Resolution.r_1080p.value, + options=Resolution, + default=Resolution.r_1080p, tooltip="Supported values may vary by model & duration", optional=True, ), comfy_io.Combo.Input( "movement_amplitude", - options=[model.value for model in MovementAmplitude], + options=MovementAmplitude, default=MovementAmplitude.auto.value, tooltip="The movement amplitude of objects in the frame", optional=True, @@ -388,8 +388,8 @@ class ViduReferenceVideoNode(comfy_io.ComfyNode): inputs=[ comfy_io.Combo.Input( "model", - options=[model.value for model in VideoModelName], - default=VideoModelName.vidu_q1.value, + options=VideoModelName, + default=VideoModelName.vidu_q1, tooltip="Model name", ), comfy_io.Image.Input( @@ -424,8 +424,8 @@ class ViduReferenceVideoNode(comfy_io.ComfyNode): ), comfy_io.Combo.Input( "aspect_ratio", - options=[model.value for model in AspectRatio], - default=AspectRatio.r_16_9.value, + options=AspectRatio, + default=AspectRatio.r_16_9, tooltip="The aspect ratio of the output video", optional=True, ),