From 7601e89255cde24667d3b4e6022f1385d901748b Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Wed, 19 Nov 2025 17:17:15 -0800 Subject: [PATCH 1/8] Fix workflow name. (#10806) --- .github/workflows/release-stable-all.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/release-stable-all.yml b/.github/workflows/release-stable-all.yml index f7de3a7c3..9274b4170 100644 --- a/.github/workflows/release-stable-all.yml +++ b/.github/workflows/release-stable-all.yml @@ -14,7 +14,7 @@ jobs: contents: "write" packages: "write" pull-requests: "read" - name: "Release NVIDIA Default (cu129)" + name: "Release NVIDIA Default (cu130)" uses: ./.github/workflows/stable-release.yml with: git_tag: ${{ inputs.git_tag }} From 394348f5caaa062eac11a57e2997aacccd4246eb Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Thu, 20 Nov 2025 03:44:04 +0200 Subject: [PATCH 2/8] feat(api-nodes): add Topaz API nodes (#10755) --- comfy_api_nodes/apis/topaz_api.py | 133 ++++++++++ comfy_api_nodes/nodes_topaz.py | 421 ++++++++++++++++++++++++++++++ comfy_api_nodes/util/client.py | 9 +- nodes.py | 1 + 4 files changed, 560 insertions(+), 4 deletions(-) create mode 100644 comfy_api_nodes/apis/topaz_api.py create mode 100644 comfy_api_nodes/nodes_topaz.py diff --git a/comfy_api_nodes/apis/topaz_api.py b/comfy_api_nodes/apis/topaz_api.py new file mode 100644 index 000000000..4d9e62e72 --- /dev/null +++ b/comfy_api_nodes/apis/topaz_api.py @@ -0,0 +1,133 @@ +from typing import Optional, Union + +from pydantic import BaseModel, Field + + +class ImageEnhanceRequest(BaseModel): + model: str = Field("Reimagine") + output_format: str = Field("jpeg") + subject_detection: str = Field("All") + face_enhancement: bool = Field(True) + face_enhancement_creativity: float = Field(0, description="Is ignored if face_enhancement is false") + face_enhancement_strength: float = Field(0.8, description="Is ignored if face_enhancement is false") + source_url: str = Field(...) + output_width: Optional[int] = Field(None) + output_height: Optional[int] = Field(None) + crop_to_fill: bool = Field(False) + prompt: Optional[str] = Field(None, description="Text prompt for creative upscaling guidance") + creativity: int = Field(3, description="Creativity settings range from 1 to 9") + face_preservation: str = Field("true", description="To preserve the identity of characters") + color_preservation: str = Field("true", description="To preserve the original color") + + +class ImageAsyncTaskResponse(BaseModel): + process_id: str = Field(...) + + +class ImageStatusResponse(BaseModel): + process_id: str = Field(...) + status: str = Field(...) + progress: Optional[int] = Field(None) + credits: int = Field(...) + + +class ImageDownloadResponse(BaseModel): + download_url: str = Field(...) + expiry: int = Field(...) + + +class Resolution(BaseModel): + width: int = Field(...) + height: int = Field(...) + + +class CreateCreateVideoRequestSource(BaseModel): + container: str = Field(...) + size: int = Field(..., description="Size of the video file in bytes") + duration: int = Field(..., description="Duration of the video file in seconds") + frameCount: int = Field(..., description="Total number of frames in the video") + frameRate: int = Field(...) + resolution: Resolution = Field(...) + + +class VideoFrameInterpolationFilter(BaseModel): + model: str = Field(...) + slowmo: Optional[int] = Field(None) + fps: int = Field(...) + duplicate: bool = Field(...) + duplicate_threshold: float = Field(...) + + +class VideoEnhancementFilter(BaseModel): + model: str = Field(...) + auto: Optional[str] = Field(None, description="Auto, Manual, Relative") + focusFixLevel: Optional[str] = Field(None, description="Downscales video input for correction of blurred subjects") + compression: Optional[float] = Field(None, description="Strength of compression recovery") + details: Optional[float] = Field(None, description="Amount of detail reconstruction") + prenoise: Optional[float] = Field(None, description="Amount of noise to add to input to reduce over-smoothing") + noise: Optional[float] = Field(None, description="Amount of noise reduction") + halo: Optional[float] = Field(None, description="Amount of halo reduction") + preblur: Optional[float] = Field(None, description="Anti-aliasing and deblurring strength") + blur: Optional[float] = Field(None, description="Amount of sharpness applied") + grain: Optional[float] = Field(None, description="Grain after AI model processing") + grainSize: Optional[float] = Field(None, description="Size of generated grain") + recoverOriginalDetailValue: Optional[float] = Field(None, description="Source details into the output video") + creativity: Optional[str] = Field(None, description="Creativity level(high, low) for slc-1 only") + isOptimizedMode: Optional[bool] = Field(None, description="Set to true for Starlight Creative (slc-1) only") + + +class OutputInformationVideo(BaseModel): + resolution: Resolution = Field(...) + frameRate: int = Field(...) + audioCodec: Optional[str] = Field(..., description="Required if audioTransfer is Copy or Convert") + audioTransfer: str = Field(..., description="Copy, Convert, None") + dynamicCompressionLevel: str = Field(..., description="Low, Mid, High") + + +class Overrides(BaseModel): + isPaidDiffusion: bool = Field(True) + + +class CreateVideoRequest(BaseModel): + source: CreateCreateVideoRequestSource = Field(...) + filters: list[Union[VideoFrameInterpolationFilter, VideoEnhancementFilter]] = Field(...) + output: OutputInformationVideo = Field(...) + overrides: Overrides = Field(Overrides(isPaidDiffusion=True)) + + +class CreateVideoResponse(BaseModel): + requestId: str = Field(...) + + +class VideoAcceptResponse(BaseModel): + uploadId: str = Field(...) + urls: list[str] = Field(...) + + +class VideoCompleteUploadRequestPart(BaseModel): + partNum: int = Field(...) + eTag: str = Field(...) + + +class VideoCompleteUploadRequest(BaseModel): + uploadResults: list[VideoCompleteUploadRequestPart] = Field(...) + + +class VideoCompleteUploadResponse(BaseModel): + message: str = Field(..., description="Confirmation message") + + +class VideoStatusResponseEstimates(BaseModel): + cost: list[int] = Field(...) + + +class VideoStatusResponseDownloadUrl(BaseModel): + url: str = Field(...) + + +class VideoStatusResponse(BaseModel): + status: str = Field(...) + estimates: Optional[VideoStatusResponseEstimates] = Field(None) + progress: Optional[float] = Field(None) + message: Optional[str] = Field("") + download: Optional[VideoStatusResponseDownloadUrl] = Field(None) diff --git a/comfy_api_nodes/nodes_topaz.py b/comfy_api_nodes/nodes_topaz.py new file mode 100644 index 000000000..79c7bf43d --- /dev/null +++ b/comfy_api_nodes/nodes_topaz.py @@ -0,0 +1,421 @@ +import builtins +from io import BytesIO + +import aiohttp +import torch +from typing_extensions import override + +from comfy_api.input.video_types import VideoInput +from comfy_api.latest import IO, ComfyExtension +from comfy_api_nodes.apis import topaz_api +from comfy_api_nodes.util import ( + ApiEndpoint, + download_url_to_image_tensor, + download_url_to_video_output, + get_fs_object_size, + get_number_of_images, + poll_op, + sync_op, + upload_images_to_comfyapi, + validate_container_format_is_mp4, +) + +UPSCALER_MODELS_MAP = { + "Starlight (Astra) Fast": "slf-1", + "Starlight (Astra) Creative": "slc-1", +} +UPSCALER_VALUES_MAP = { + "FullHD (1080p)": 1920, + "4K (2160p)": 3840, +} + + +class TopazImageEnhance(IO.ComfyNode): + @classmethod + def define_schema(cls): + return IO.Schema( + node_id="TopazImageEnhance", + display_name="Topaz Image Enhance", + category="api node/image/Topaz", + description="Industry-standard upscaling and image enhancement.", + inputs=[ + IO.Combo.Input("model", options=["Reimagine"]), + IO.Image.Input("image"), + IO.String.Input( + "prompt", + multiline=True, + default="", + tooltip="Optional text prompt for creative upscaling guidance.", + optional=True, + ), + IO.Combo.Input( + "subject_detection", + options=["All", "Foreground", "Background"], + optional=True, + ), + IO.Boolean.Input( + "face_enhancement", + default=True, + optional=True, + tooltip="Enhance faces (if present) during processing.", + ), + IO.Float.Input( + "face_enhancement_creativity", + default=0.0, + min=0.0, + max=1.0, + step=0.01, + display_mode=IO.NumberDisplay.number, + optional=True, + tooltip="Set the creativity level for face enhancement.", + ), + IO.Float.Input( + "face_enhancement_strength", + default=1.0, + min=0.0, + max=1.0, + step=0.01, + display_mode=IO.NumberDisplay.number, + optional=True, + tooltip="Controls how sharp enhanced faces are relative to the background.", + ), + IO.Boolean.Input( + "crop_to_fill", + default=False, + optional=True, + tooltip="By default, the image is letterboxed when the output aspect ratio differs. " + "Enable to crop the image to fill the output dimensions.", + ), + IO.Int.Input( + "output_width", + default=0, + min=0, + max=32000, + step=1, + display_mode=IO.NumberDisplay.number, + optional=True, + tooltip="Zero value means to calculate automatically (usually it will be original size or output_height if specified).", + ), + IO.Int.Input( + "output_height", + default=0, + min=0, + max=32000, + step=1, + display_mode=IO.NumberDisplay.number, + optional=True, + tooltip="Zero value means to output in the same height as original or output width.", + ), + IO.Int.Input( + "creativity", + default=3, + min=1, + max=9, + step=1, + display_mode=IO.NumberDisplay.slider, + optional=True, + ), + IO.Boolean.Input( + "face_preservation", + default=True, + optional=True, + tooltip="Preserve subjects' facial identity.", + ), + IO.Boolean.Input( + "color_preservation", + default=True, + optional=True, + tooltip="Preserve the original colors.", + ), + ], + outputs=[ + IO.Image.Output(), + ], + hidden=[ + IO.Hidden.auth_token_comfy_org, + IO.Hidden.api_key_comfy_org, + IO.Hidden.unique_id, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, + model: str, + image: torch.Tensor, + prompt: str = "", + subject_detection: str = "All", + face_enhancement: bool = True, + face_enhancement_creativity: float = 1.0, + face_enhancement_strength: float = 0.8, + crop_to_fill: bool = False, + output_width: int = 0, + output_height: int = 0, + creativity: int = 3, + face_preservation: bool = True, + color_preservation: bool = True, + ) -> IO.NodeOutput: + if get_number_of_images(image) != 1: + raise ValueError("Only one input image is supported.") + download_url = await upload_images_to_comfyapi(cls, image, max_images=1, mime_type="image/png") + initial_response = await sync_op( + cls, + ApiEndpoint(path="/proxy/topaz/image/v1/enhance-gen/async", method="POST"), + response_model=topaz_api.ImageAsyncTaskResponse, + data=topaz_api.ImageEnhanceRequest( + model=model, + prompt=prompt, + subject_detection=subject_detection, + face_enhancement=face_enhancement, + face_enhancement_creativity=face_enhancement_creativity, + face_enhancement_strength=face_enhancement_strength, + crop_to_fill=crop_to_fill, + output_width=output_width if output_width else None, + output_height=output_height if output_height else None, + creativity=creativity, + face_preservation=str(face_preservation).lower(), + color_preservation=str(color_preservation).lower(), + source_url=download_url[0], + output_format="png", + ), + content_type="multipart/form-data", + ) + + await poll_op( + cls, + poll_endpoint=ApiEndpoint(path=f"/proxy/topaz/image/v1/status/{initial_response.process_id}"), + response_model=topaz_api.ImageStatusResponse, + status_extractor=lambda x: x.status, + progress_extractor=lambda x: getattr(x, "progress", 0), + price_extractor=lambda x: x.credits * 0.08, + poll_interval=8.0, + max_poll_attempts=160, + estimated_duration=60, + ) + + results = await sync_op( + cls, + ApiEndpoint(path=f"/proxy/topaz/image/v1/download/{initial_response.process_id}"), + response_model=topaz_api.ImageDownloadResponse, + monitor_progress=False, + ) + return IO.NodeOutput(await download_url_to_image_tensor(results.download_url)) + + +class TopazVideoEnhance(IO.ComfyNode): + @classmethod + def define_schema(cls): + return IO.Schema( + node_id="TopazVideoEnhance", + display_name="Topaz Video Enhance", + category="api node/video/Topaz", + description="Breathe new life into video with powerful upscaling and recovery technology.", + inputs=[ + IO.Video.Input("video"), + IO.Boolean.Input("upscaler_enabled", default=True), + IO.Combo.Input("upscaler_model", options=list(UPSCALER_MODELS_MAP.keys())), + IO.Combo.Input("upscaler_resolution", options=list(UPSCALER_VALUES_MAP.keys())), + IO.Combo.Input( + "upscaler_creativity", + options=["low", "middle", "high"], + default="low", + tooltip="Creativity level (applies only to Starlight (Astra) Creative).", + optional=True, + ), + IO.Boolean.Input("interpolation_enabled", default=False, optional=True), + IO.Combo.Input("interpolation_model", options=["apo-8"], default="apo-8", optional=True), + IO.Int.Input( + "interpolation_slowmo", + default=1, + min=1, + max=16, + display_mode=IO.NumberDisplay.number, + tooltip="Slow-motion factor applied to the input video. " + "For example, 2 makes the output twice as slow and doubles the duration.", + optional=True, + ), + IO.Int.Input( + "interpolation_frame_rate", + default=60, + min=15, + max=240, + display_mode=IO.NumberDisplay.number, + tooltip="Output frame rate.", + optional=True, + ), + IO.Boolean.Input( + "interpolation_duplicate", + default=False, + tooltip="Analyze the input for duplicate frames and remove them.", + optional=True, + ), + IO.Float.Input( + "interpolation_duplicate_threshold", + default=0.01, + min=0.001, + max=0.1, + step=0.001, + display_mode=IO.NumberDisplay.number, + tooltip="Detection sensitivity for duplicate frames.", + optional=True, + ), + IO.Combo.Input( + "dynamic_compression_level", + options=["Low", "Mid", "High"], + default="Low", + tooltip="CQP level.", + optional=True, + ), + ], + outputs=[ + IO.Video.Output(), + ], + hidden=[ + IO.Hidden.auth_token_comfy_org, + IO.Hidden.api_key_comfy_org, + IO.Hidden.unique_id, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, + video: VideoInput, + upscaler_enabled: bool, + upscaler_model: str, + upscaler_resolution: str, + upscaler_creativity: str = "low", + interpolation_enabled: bool = False, + interpolation_model: str = "apo-8", + interpolation_slowmo: int = 1, + interpolation_frame_rate: int = 60, + interpolation_duplicate: bool = False, + interpolation_duplicate_threshold: float = 0.01, + dynamic_compression_level: str = "Low", + ) -> IO.NodeOutput: + if upscaler_enabled is False and interpolation_enabled is False: + raise ValueError("There is nothing to do: both upscaling and interpolation are disabled.") + src_width, src_height = video.get_dimensions() + video_components = video.get_components() + src_frame_rate = int(video_components.frame_rate) + duration_sec = video.get_duration() + estimated_frames = int(duration_sec * src_frame_rate) + validate_container_format_is_mp4(video) + src_video_stream = video.get_stream_source() + target_width = src_width + target_height = src_height + target_frame_rate = src_frame_rate + filters = [] + if upscaler_enabled: + target_width = UPSCALER_VALUES_MAP[upscaler_resolution] + target_height = UPSCALER_VALUES_MAP[upscaler_resolution] + filters.append( + topaz_api.VideoEnhancementFilter( + model=UPSCALER_MODELS_MAP[upscaler_model], + creativity=(upscaler_creativity if UPSCALER_MODELS_MAP[upscaler_model] == "slc-1" else None), + isOptimizedMode=(True if UPSCALER_MODELS_MAP[upscaler_model] == "slc-1" else None), + ), + ) + if interpolation_enabled: + target_frame_rate = interpolation_frame_rate + filters.append( + topaz_api.VideoFrameInterpolationFilter( + model=interpolation_model, + slowmo=interpolation_slowmo, + fps=interpolation_frame_rate, + duplicate=interpolation_duplicate, + duplicate_threshold=interpolation_duplicate_threshold, + ), + ) + initial_res = await sync_op( + cls, + ApiEndpoint(path="/proxy/topaz/video/", method="POST"), + response_model=topaz_api.CreateVideoResponse, + data=topaz_api.CreateVideoRequest( + source=topaz_api.CreateCreateVideoRequestSource( + container="mp4", + size=get_fs_object_size(src_video_stream), + duration=int(duration_sec), + frameCount=estimated_frames, + frameRate=src_frame_rate, + resolution=topaz_api.Resolution(width=src_width, height=src_height), + ), + filters=filters, + output=topaz_api.OutputInformationVideo( + resolution=topaz_api.Resolution(width=target_width, height=target_height), + frameRate=target_frame_rate, + audioCodec="AAC", + audioTransfer="Copy", + dynamicCompressionLevel=dynamic_compression_level, + ), + ), + wait_label="Creating task", + final_label_on_success="Task created", + ) + upload_res = await sync_op( + cls, + ApiEndpoint( + path=f"/proxy/topaz/video/{initial_res.requestId}/accept", + method="PATCH", + ), + response_model=topaz_api.VideoAcceptResponse, + wait_label="Preparing upload", + final_label_on_success="Upload started", + ) + if len(upload_res.urls) > 1: + raise NotImplementedError( + "Large files are not currently supported. Please open an issue in the ComfyUI repository." + ) + async with aiohttp.ClientSession(headers={"Content-Type": "video/mp4"}) as session: + if isinstance(src_video_stream, BytesIO): + src_video_stream.seek(0) + async with session.put(upload_res.urls[0], data=src_video_stream, raise_for_status=True) as res: + upload_etag = res.headers["Etag"] + else: + with builtins.open(src_video_stream, "rb") as video_file: + async with session.put(upload_res.urls[0], data=video_file, raise_for_status=True) as res: + upload_etag = res.headers["Etag"] + await sync_op( + cls, + ApiEndpoint( + path=f"/proxy/topaz/video/{initial_res.requestId}/complete-upload", + method="PATCH", + ), + response_model=topaz_api.VideoCompleteUploadResponse, + data=topaz_api.VideoCompleteUploadRequest( + uploadResults=[ + topaz_api.VideoCompleteUploadRequestPart( + partNum=1, + eTag=upload_etag, + ), + ], + ), + wait_label="Finalizing upload", + final_label_on_success="Upload completed", + ) + final_response = await poll_op( + cls, + ApiEndpoint(path=f"/proxy/topaz/video/{initial_res.requestId}/status"), + response_model=topaz_api.VideoStatusResponse, + status_extractor=lambda x: x.status, + progress_extractor=lambda x: getattr(x, "progress", 0), + price_extractor=lambda x: (x.estimates.cost[0] * 0.08 if x.estimates and x.estimates.cost[0] else None), + poll_interval=10.0, + max_poll_attempts=320, + ) + return IO.NodeOutput(await download_url_to_video_output(final_response.download.url)) + + +class TopazExtension(ComfyExtension): + @override + async def get_node_list(self) -> list[type[IO.ComfyNode]]: + return [ + TopazImageEnhance, + TopazVideoEnhance, + ] + + +async def comfy_entrypoint() -> TopazExtension: + return TopazExtension() diff --git a/comfy_api_nodes/util/client.py b/comfy_api_nodes/util/client.py index 2d5dcd648..ad6e3c0d0 100644 --- a/comfy_api_nodes/util/client.py +++ b/comfy_api_nodes/util/client.py @@ -77,9 +77,9 @@ class _PollUIState: _RETRY_STATUS = {408, 429, 500, 502, 503, 504} -COMPLETED_STATUSES = ["succeeded", "succeed", "success", "completed", "finished", "done"] -FAILED_STATUSES = ["cancelled", "canceled", "fail", "failed", "error"] -QUEUED_STATUSES = ["created", "queued", "queueing", "submitted"] +COMPLETED_STATUSES = ["succeeded", "succeed", "success", "completed", "finished", "done", "complete"] +FAILED_STATUSES = ["cancelled", "canceled", "canceling", "fail", "failed", "error"] +QUEUED_STATUSES = ["created", "queued", "queueing", "submitted", "initializing"] async def sync_op( @@ -424,7 +424,8 @@ def _display_text( if status: display_lines.append(f"Status: {status.capitalize() if isinstance(status, str) else status}") if price is not None: - display_lines.append(f"Price: ${float(price):,.4f}") + p = f"{float(price):,.4f}".rstrip("0").rstrip(".") + display_lines.append(f"Price: ${p}") if text is not None: display_lines.append(text) if display_lines: diff --git a/nodes.py b/nodes.py index f6aeedc78..ac14e39a7 100644 --- a/nodes.py +++ b/nodes.py @@ -2359,6 +2359,7 @@ async def init_builtin_api_nodes(): "nodes_pika.py", "nodes_runway.py", "nodes_sora.py", + "nodes_topaz.py", "nodes_tripo.py", "nodes_moonvalley.py", "nodes_rodin.py", From cb96d4d18c78ee09d5fd70954ffcb4ad2c7f0d7a Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Wed, 19 Nov 2025 20:56:23 -0800 Subject: [PATCH 3/8] Disable workaround on newer cudnn. (#10807) --- comfy/ops.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/comfy/ops.py b/comfy/ops.py index 2a90a5ba2..640622fd1 100644 --- a/comfy/ops.py +++ b/comfy/ops.py @@ -58,7 +58,8 @@ except (ModuleNotFoundError, TypeError): NVIDIA_MEMORY_CONV_BUG_WORKAROUND = False try: if comfy.model_management.is_nvidia(): - if torch.backends.cudnn.version() >= 91002 and comfy.model_management.torch_version_numeric >= (2, 9) and comfy.model_management.torch_version_numeric <= (2, 10): + cudnn_version = torch.backends.cudnn.version() + if (cudnn_version >= 91002 and cudnn_version < 91500) and comfy.model_management.torch_version_numeric >= (2, 9) and comfy.model_management.torch_version_numeric <= (2, 10): #TODO: change upper bound version once it's fixed' NVIDIA_MEMORY_CONV_BUG_WORKAROUND = True logging.info("working around nvidia conv3d memory bug.") From 87b0359392219841c2214e1eb06678840cae470e Mon Sep 17 00:00:00 2001 From: Christian Byrne Date: Wed, 19 Nov 2025 22:36:56 -0800 Subject: [PATCH 4/8] Update server templates handler to use new multi-package distribution (comfyui-workflow-templates versions >=0.3) (#10791) * update templates for monorepo * refactor --- app/frontend_management.py | 67 ++++++++++++++++++++++++++++++++++++-- requirements.txt | 2 +- server.py | 32 ++++++++++++++---- 3 files changed, 92 insertions(+), 9 deletions(-) diff --git a/app/frontend_management.py b/app/frontend_management.py index cce0c117d..bdaa85812 100644 --- a/app/frontend_management.py +++ b/app/frontend_management.py @@ -10,7 +10,8 @@ import importlib from dataclasses import dataclass from functools import cached_property from pathlib import Path -from typing import TypedDict, Optional +from typing import Dict, TypedDict, Optional +from aiohttp import web from importlib.metadata import version import requests @@ -257,7 +258,54 @@ comfyui-frontend-package is not installed. sys.exit(-1) @classmethod - def templates_path(cls) -> str: + def template_asset_map(cls) -> Optional[Dict[str, str]]: + """Return a mapping of template asset names to their absolute paths.""" + try: + from comfyui_workflow_templates import ( + get_asset_path, + iter_templates, + ) + except ImportError: + logging.error( + f""" +********** ERROR *********** + +comfyui-workflow-templates is not installed. + +{frontend_install_warning_message()} + +********** ERROR *********** +""".strip() + ) + return None + + try: + template_entries = list(iter_templates()) + except Exception as exc: + logging.error(f"Failed to enumerate workflow templates: {exc}") + return None + + asset_map: Dict[str, str] = {} + try: + for entry in template_entries: + for asset in entry.assets: + asset_map[asset.filename] = get_asset_path( + entry.template_id, asset.filename + ) + except Exception as exc: + logging.error(f"Failed to resolve template asset paths: {exc}") + return None + + if not asset_map: + logging.error("No workflow template assets found. Did the packages install correctly?") + return None + + return asset_map + + + @classmethod + def legacy_templates_path(cls) -> Optional[str]: + """Return the legacy templates directory shipped inside the meta package.""" try: import comfyui_workflow_templates @@ -276,6 +324,7 @@ comfyui-workflow-templates is not installed. ********** ERROR *********** """.strip() ) + return None @classmethod def embedded_docs_path(cls) -> str: @@ -392,3 +441,17 @@ comfyui-workflow-templates is not installed. logging.info("Falling back to the default frontend.") check_frontend_version() return cls.default_frontend_path() + @classmethod + def template_asset_handler(cls): + assets = cls.template_asset_map() + if not assets: + return None + + async def serve_template(request: web.Request) -> web.StreamResponse: + rel_path = request.match_info.get("path", "") + target = assets.get(rel_path) + if target is None: + raise web.HTTPNotFound() + return web.FileResponse(target) + + return serve_template diff --git a/requirements.txt b/requirements.txt index 249c36dee..36c39f338 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,5 +1,5 @@ comfyui-frontend-package==1.28.8 -comfyui-workflow-templates==0.2.11 +comfyui-workflow-templates==0.3.1 comfyui-embedded-docs==0.3.1 torch torchsde diff --git a/server.py b/server.py index d059d3dc9..d9d5c491f 100644 --- a/server.py +++ b/server.py @@ -30,7 +30,7 @@ import comfy.model_management from comfy_api import feature_flags import node_helpers from comfyui_version import __version__ -from app.frontend_management import FrontendManager +from app.frontend_management import FrontendManager, parse_version from comfy_api.internal import _ComfyNodeInternal from app.user_manager import UserManager @@ -849,11 +849,31 @@ class PromptServer(): for name, dir in nodes.EXTENSION_WEB_DIRS.items(): self.app.add_routes([web.static('/extensions/' + name, dir)]) - workflow_templates_path = FrontendManager.templates_path() - if workflow_templates_path: - self.app.add_routes([ - web.static('/templates', workflow_templates_path) - ]) + installed_templates_version = FrontendManager.get_installed_templates_version() + use_legacy_templates = True + if installed_templates_version: + try: + use_legacy_templates = ( + parse_version(installed_templates_version) + < parse_version("0.3.0") + ) + except Exception as exc: + logging.warning( + "Unable to parse templates version '%s': %s", + installed_templates_version, + exc, + ) + + if use_legacy_templates: + workflow_templates_path = FrontendManager.legacy_templates_path() + if workflow_templates_path: + self.app.add_routes([ + web.static('/templates', workflow_templates_path) + ]) + else: + handler = FrontendManager.template_asset_handler() + if handler: + self.app.router.add_get("/templates/{path:.*}", handler) # Serve embedded documentation from the package embedded_docs_path = FrontendManager.embedded_docs_path() From f5e66d5e47271253edad5c4eddd817b0d6a23340 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Thu, 20 Nov 2025 12:08:03 -0800 Subject: [PATCH 5/8] Fix ImageBatch with different channel count. (#10815) --- nodes.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/nodes.py b/nodes.py index ac14e39a7..75e820e66 100644 --- a/nodes.py +++ b/nodes.py @@ -1852,6 +1852,10 @@ class ImageBatch: CATEGORY = "image" def batch(self, image1, image2): + if image1.shape[-1] != image2.shape[-1]: + channels = min(image1.shape[-1], image2.shape[-1]) + image1 = image1[..., :channels] + image2 = image2[..., :channels] if image1.shape[1:] != image2.shape[1:]: image2 = comfy.utils.common_upscale(image2.movedim(-1,1), image1.shape[2], image1.shape[1], "bilinear", "center").movedim(1,-1) s = torch.cat((image1, image2), dim=0) From 9e00ce5b76ec04be37375310512a443605b95077 Mon Sep 17 00:00:00 2001 From: Jedrzej Kosinski Date: Thu, 20 Nov 2025 14:42:46 -0800 Subject: [PATCH 6/8] Make Batch Images node add alpha channel when one of the inputs has it (#10816) * When one Batch Image input has alpha and one does not, add empty alpha channel * Use torch.nn.functional.pad --- nodes.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/nodes.py b/nodes.py index 75e820e66..030371633 100644 --- a/nodes.py +++ b/nodes.py @@ -1853,9 +1853,10 @@ class ImageBatch: def batch(self, image1, image2): if image1.shape[-1] != image2.shape[-1]: - channels = min(image1.shape[-1], image2.shape[-1]) - image1 = image1[..., :channels] - image2 = image2[..., :channels] + if image1.shape[-1] > image2.shape[-1]: + image2 = torch.nn.functional.pad(image2, (0,1), mode='constant', value=1.0) + else: + image1 = torch.nn.functional.pad(image1, (0,1), mode='constant', value=1.0) if image1.shape[1:] != image2.shape[1:]: image2 = comfy.utils.common_upscale(image2.movedim(-1,1), image1.shape[2], image1.shape[1], "bilinear", "center").movedim(1,-1) s = torch.cat((image1, image2), dim=0) From 7b8389578e88dcd13b1cf6aea5404047298c9183 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 21 Nov 2025 02:17:47 +0200 Subject: [PATCH 7/8] feat(api-nodes): add Nano Banana Pro (#10814) * feat(api-nodes): add Nano Banana Pro * frontend bump to 1.28.9 --- comfy_api_nodes/apis/gemini_api.py | 5 +- comfy_api_nodes/nodes_gemini.py | 205 ++++++++++++++++++++++++++++- comfy_api_nodes/util/client.py | 13 +- requirements.txt | 2 +- 4 files changed, 215 insertions(+), 10 deletions(-) diff --git a/comfy_api_nodes/apis/gemini_api.py b/comfy_api_nodes/apis/gemini_api.py index f63e02693..710f173f1 100644 --- a/comfy_api_nodes/apis/gemini_api.py +++ b/comfy_api_nodes/apis/gemini_api.py @@ -68,7 +68,7 @@ class GeminiTextPart(BaseModel): class GeminiContent(BaseModel): - parts: list[GeminiPart] = Field(...) + parts: list[GeminiPart] = Field([]) role: GeminiRole = Field(..., examples=["user"]) @@ -120,7 +120,7 @@ class GeminiGenerationConfig(BaseModel): class GeminiImageConfig(BaseModel): aspectRatio: str | None = Field(None) - resolution: str | None = Field(None) + imageSize: str | None = Field(None) class GeminiImageGenerationConfig(GeminiGenerationConfig): @@ -227,3 +227,4 @@ class GeminiGenerateContentResponse(BaseModel): candidates: list[GeminiCandidate] | None = Field(None) promptFeedback: GeminiPromptFeedback | None = Field(None) usageMetadata: GeminiUsageMetadata | None = Field(None) + modelVersion: str | None = Field(None) diff --git a/comfy_api_nodes/nodes_gemini.py b/comfy_api_nodes/nodes_gemini.py index 6e746eebd..be752c885 100644 --- a/comfy_api_nodes/nodes_gemini.py +++ b/comfy_api_nodes/nodes_gemini.py @@ -29,11 +29,13 @@ from comfy_api_nodes.apis.gemini_api import ( GeminiMimeType, GeminiPart, GeminiRole, + Modality, ) from comfy_api_nodes.util import ( ApiEndpoint, audio_to_base64_string, bytesio_to_image_tensor, + get_number_of_images, sync_op, tensor_to_base64_string, validate_string, @@ -147,6 +149,49 @@ def get_image_from_response(response: GeminiGenerateContentResponse) -> torch.Te return torch.cat(image_tensors, dim=0) +def calculate_tokens_price(response: GeminiGenerateContentResponse) -> float | None: + if not response.modelVersion: + return None + # Define prices (Cost per 1,000,000 tokens), see https://cloud.google.com/vertex-ai/generative-ai/pricing + if response.modelVersion in ("gemini-2.5-pro-preview-05-06", "gemini-2.5-pro"): + input_tokens_price = 1.25 + output_text_tokens_price = 10.0 + output_image_tokens_price = 0.0 + elif response.modelVersion in ( + "gemini-2.5-flash-preview-04-17", + "gemini-2.5-flash", + ): + input_tokens_price = 0.30 + output_text_tokens_price = 2.50 + output_image_tokens_price = 0.0 + elif response.modelVersion in ( + "gemini-2.5-flash-image-preview", + "gemini-2.5-flash-image", + ): + input_tokens_price = 0.30 + output_text_tokens_price = 2.50 + output_image_tokens_price = 30.0 + elif response.modelVersion == "gemini-3-pro-preview": + input_tokens_price = 2 + output_text_tokens_price = 12.0 + output_image_tokens_price = 0.0 + elif response.modelVersion == "gemini-3-pro-image-preview": + input_tokens_price = 2 + output_text_tokens_price = 12.0 + output_image_tokens_price = 120.0 + else: + return None + final_price = response.usageMetadata.promptTokenCount * input_tokens_price + for i in response.usageMetadata.candidatesTokensDetails: + if i.modality == Modality.IMAGE: + final_price += output_image_tokens_price * i.tokenCount # for Nano Banana models + else: + final_price += output_text_tokens_price * i.tokenCount + if response.usageMetadata.thoughtsTokenCount: + final_price += output_text_tokens_price * response.usageMetadata.thoughtsTokenCount + return final_price / 1_000_000.0 + + class GeminiNode(IO.ComfyNode): """ Node to generate text responses from a Gemini model. @@ -314,6 +359,7 @@ class GeminiNode(IO.ComfyNode): ] ), response_model=GeminiGenerateContentResponse, + price_extractor=calculate_tokens_price, ) output_text = get_text_from_response(response) @@ -476,6 +522,13 @@ class GeminiImage(IO.ComfyNode): "or otherwise generates 1:1 squares.", optional=True, ), + IO.Combo.Input( + "response_modalities", + options=["IMAGE+TEXT", "IMAGE"], + tooltip="Choose 'IMAGE' for image-only output, or " + "'IMAGE+TEXT' to return both the generated image and a text response.", + optional=True, + ), ], outputs=[ IO.Image.Output(), @@ -498,6 +551,7 @@ class GeminiImage(IO.ComfyNode): images: torch.Tensor | None = None, files: list[GeminiPart] | None = None, aspect_ratio: str = "auto", + response_modalities: str = "IMAGE+TEXT", ) -> IO.NodeOutput: validate_string(prompt, strip_whitespace=True, min_length=1) parts: list[GeminiPart] = [GeminiPart(text=prompt)] @@ -520,17 +574,16 @@ class GeminiImage(IO.ComfyNode): GeminiContent(role=GeminiRole.user, parts=parts), ], generationConfig=GeminiImageGenerationConfig( - responseModalities=["TEXT", "IMAGE"], + responseModalities=(["IMAGE"] if response_modalities == "IMAGE" else ["TEXT", "IMAGE"]), imageConfig=None if aspect_ratio == "auto" else image_config, ), ), response_model=GeminiGenerateContentResponse, + price_extractor=calculate_tokens_price, ) - output_image = get_image_from_response(response) output_text = get_text_from_response(response) if output_text: - # Not a true chat history like the OpenAI Chat node. It is emulated so the frontend can show a copy button. render_spec = { "node_id": cls.hidden.unique_id, "component": "ChatHistoryWidget", @@ -551,9 +604,150 @@ class GeminiImage(IO.ComfyNode): "display_component", render_spec, ) + return IO.NodeOutput(get_image_from_response(response), output_text) - output_text = output_text or "Empty response from Gemini model..." - return IO.NodeOutput(output_image, output_text) + +class GeminiImage2(IO.ComfyNode): + + @classmethod + def define_schema(cls): + return IO.Schema( + node_id="GeminiImage2Node", + display_name="Nano Banana Pro (Google Gemini Image)", + category="api node/image/Gemini", + description="Generate or edit images synchronously via Google Vertex API.", + inputs=[ + IO.String.Input( + "prompt", + multiline=True, + tooltip="Text prompt describing the image to generate or the edits to apply. " + "Include any constraints, styles, or details the model should follow.", + default="", + ), + IO.Combo.Input( + "model", + options=["gemini-3-pro-image-preview"], + ), + IO.Int.Input( + "seed", + default=42, + min=0, + max=0xFFFFFFFFFFFFFFFF, + control_after_generate=True, + tooltip="When the seed is fixed to a specific value, the model makes a best effort to provide " + "the same response for repeated requests. Deterministic output isn't guaranteed. " + "Also, changing the model or parameter settings, such as the temperature, " + "can cause variations in the response even when you use the same seed value. " + "By default, a random seed value is used.", + ), + IO.Combo.Input( + "aspect_ratio", + options=["auto", "1:1", "2:3", "3:2", "3:4", "4:3", "4:5", "5:4", "9:16", "16:9", "21:9"], + default="auto", + tooltip="If set to 'auto', matches your input image's aspect ratio; " + "if no image is provided, generates a 1:1 square.", + ), + IO.Combo.Input( + "resolution", + options=["1K", "2K", "4K"], + tooltip="Target output resolution. For 2K/4K the native Gemini upscaler is used.", + ), + IO.Combo.Input( + "response_modalities", + options=["IMAGE+TEXT", "IMAGE"], + tooltip="Choose 'IMAGE' for image-only output, or " + "'IMAGE+TEXT' to return both the generated image and a text response.", + ), + IO.Image.Input( + "images", + optional=True, + tooltip="Optional reference image(s). " + "To include multiple images, use the Batch Images node (up to 14).", + ), + IO.Custom("GEMINI_INPUT_FILES").Input( + "files", + optional=True, + tooltip="Optional file(s) to use as context for the model. " + "Accepts inputs from the Gemini Generate Content Input Files node.", + ), + ], + outputs=[ + IO.Image.Output(), + IO.String.Output(), + ], + hidden=[ + IO.Hidden.auth_token_comfy_org, + IO.Hidden.api_key_comfy_org, + IO.Hidden.unique_id, + ], + is_api_node=True, + ) + + @classmethod + async def execute( + cls, + prompt: str, + model: str, + seed: int, + aspect_ratio: str, + resolution: str, + response_modalities: str, + images: torch.Tensor | None = None, + files: list[GeminiPart] | None = None, + ) -> IO.NodeOutput: + validate_string(prompt, strip_whitespace=True, min_length=1) + + parts: list[GeminiPart] = [GeminiPart(text=prompt)] + if images is not None: + if get_number_of_images(images) > 14: + raise ValueError("The current maximum number of supported images is 14.") + parts.extend(create_image_parts(images)) + if files is not None: + parts.extend(files) + + image_config = GeminiImageConfig(imageSize=resolution) + if aspect_ratio != "auto": + image_config.aspectRatio = aspect_ratio + + response = await sync_op( + cls, + ApiEndpoint(path=f"{GEMINI_BASE_ENDPOINT}/{model}", method="POST"), + data=GeminiImageGenerateContentRequest( + contents=[ + GeminiContent(role=GeminiRole.user, parts=parts), + ], + generationConfig=GeminiImageGenerationConfig( + responseModalities=(["IMAGE"] if response_modalities == "IMAGE" else ["TEXT", "IMAGE"]), + imageConfig=image_config, + ), + ), + response_model=GeminiGenerateContentResponse, + price_extractor=calculate_tokens_price, + ) + + output_text = get_text_from_response(response) + if output_text: + render_spec = { + "node_id": cls.hidden.unique_id, + "component": "ChatHistoryWidget", + "props": { + "history": json.dumps( + [ + { + "prompt": prompt, + "response": output_text, + "response_id": str(uuid.uuid4()), + "timestamp": time.time(), + } + ] + ), + }, + } + PromptServer.instance.send_sync( + "display_component", + render_spec, + ) + return IO.NodeOutput(get_image_from_response(response), output_text) class GeminiExtension(ComfyExtension): @@ -562,6 +756,7 @@ class GeminiExtension(ComfyExtension): return [ GeminiNode, GeminiImage, + GeminiImage2, GeminiInputFiles, ] diff --git a/comfy_api_nodes/util/client.py b/comfy_api_nodes/util/client.py index ad6e3c0d0..bf01d7d36 100644 --- a/comfy_api_nodes/util/client.py +++ b/comfy_api_nodes/util/client.py @@ -63,6 +63,7 @@ class _RequestConfig: estimated_total: Optional[int] = None final_label_on_success: Optional[str] = "Completed" progress_origin_ts: Optional[float] = None + price_extractor: Optional[Callable[[dict[str, Any]], Optional[float]]] = None @dataclass @@ -87,6 +88,7 @@ async def sync_op( endpoint: ApiEndpoint, *, response_model: Type[M], + price_extractor: Optional[Callable[[M], Optional[float]]] = None, data: Optional[BaseModel] = None, files: Optional[Union[dict[str, Any], list[tuple[str, Any]]]] = None, content_type: str = "application/json", @@ -104,6 +106,7 @@ async def sync_op( raw = await sync_op_raw( cls, endpoint, + price_extractor=_wrap_model_extractor(response_model, price_extractor), data=data, files=files, content_type=content_type, @@ -175,6 +178,7 @@ async def sync_op_raw( cls: type[IO.ComfyNode], endpoint: ApiEndpoint, *, + price_extractor: Optional[Callable[[dict[str, Any]], Optional[float]]] = None, data: Optional[Union[dict[str, Any], BaseModel]] = None, files: Optional[Union[dict[str, Any], list[tuple[str, Any]]]] = None, content_type: str = "application/json", @@ -216,6 +220,7 @@ async def sync_op_raw( estimated_total=estimated_duration, final_label_on_success=final_label_on_success, progress_origin_ts=progress_origin_ts, + price_extractor=price_extractor, ) return await _request_base(cfg, expect_binary=as_binary) @@ -425,7 +430,8 @@ def _display_text( display_lines.append(f"Status: {status.capitalize() if isinstance(status, str) else status}") if price is not None: p = f"{float(price):,.4f}".rstrip("0").rstrip(".") - display_lines.append(f"Price: ${p}") + if p != "0": + display_lines.append(f"Price: ${p}") if text is not None: display_lines.append(text) if display_lines: @@ -581,6 +587,7 @@ async def _request_base(cfg: _RequestConfig, expect_binary: bool): delay = cfg.retry_delay operation_succeeded: bool = False final_elapsed_seconds: Optional[int] = None + extracted_price: Optional[float] = None while True: attempt += 1 stop_event = asyncio.Event() @@ -768,6 +775,8 @@ async def _request_base(cfg: _RequestConfig, expect_binary: bool): except json.JSONDecodeError: payload = {"_raw": text} response_content_to_log = payload if isinstance(payload, dict) else text + with contextlib.suppress(Exception): + extracted_price = cfg.price_extractor(payload) if cfg.price_extractor else None operation_succeeded = True final_elapsed_seconds = int(time.monotonic() - start_time) try: @@ -872,7 +881,7 @@ async def _request_base(cfg: _RequestConfig, expect_binary: bool): else int(time.monotonic() - start_time) ), estimated_total=cfg.estimated_total, - price=None, + price=extracted_price, is_queued=False, processing_elapsed_seconds=final_elapsed_seconds, ) diff --git a/requirements.txt b/requirements.txt index 36c39f338..8c1946f3d 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,4 +1,4 @@ -comfyui-frontend-package==1.28.8 +comfyui-frontend-package==1.28.9 comfyui-workflow-templates==0.3.1 comfyui-embedded-docs==0.3.1 torch From b75d349f25ccb702895c6f1b8af7aded63a7f7e2 Mon Sep 17 00:00:00 2001 From: Alexander Piskun <13381981+bigcat88@users.noreply.github.com> Date: Fri, 21 Nov 2025 02:33:54 +0200 Subject: [PATCH 8/8] fix(KlingLipSyncAudioToVideoNode): convert audio to mp3 format (#10811) --- comfy_api_nodes/nodes_kling.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/comfy_api_nodes/nodes_kling.py b/comfy_api_nodes/nodes_kling.py index 7b23e9cf9..36852038b 100644 --- a/comfy_api_nodes/nodes_kling.py +++ b/comfy_api_nodes/nodes_kling.py @@ -518,7 +518,9 @@ async def execute_lipsync( # Upload the audio file to Comfy API and get download URL if audio: - audio_url = await upload_audio_to_comfyapi(cls, audio) + audio_url = await upload_audio_to_comfyapi( + cls, audio, container_format="mp3", codec_name="libmp3lame", mime_type="audio/mpeg", filename="output.mp3" + ) logging.info("Uploaded audio to Comfy API. URL: %s", audio_url) else: audio_url = None