diff --git a/README.md b/README.md
index ee1024de5..786a14166 100644
--- a/README.md
+++ b/README.md
@@ -364,7 +364,7 @@ For models compatible with Iluvatar Extension for PyTorch. Here's a step-by-step
 | Flag | Description |
 |------|-------------|
 | `--enable-manager` | Enable ComfyUI-Manager |
-| `--enable-manager-legacy-ui` | Use the legacy manager UI instead of the new UI (requires `--enable-manager`) |
+| `--enable-manager-legacy-ui` | Use the legacy manager UI instead of the new UI (implies `--enable-manager`) |
 | `--disable-manager-ui` | Disable the manager UI and endpoints while keeping background features like security checks and scheduled installation completion (requires `--enable-manager`) |
 
 
diff --git a/comfy/cli_args.py b/comfy/cli_args.py
index cba0dfa34..e7ee0d5eb 100644
--- a/comfy/cli_args.py
+++ b/comfy/cli_args.py
@@ -115,6 +115,7 @@ cache_group.add_argument("--cache-ram", nargs='*', type=float, default=[], metav
 cache_group.add_argument("--cache-classic", action="store_true", help="Use the old style (aggressive) caching.")
 cache_group.add_argument("--cache-lru", type=int, default=0, help="Use LRU caching with a maximum of N node results cached. May use more RAM/VRAM.")
 cache_group.add_argument("--cache-none", action="store_true", help="Reduced RAM/VRAM usage at the expense of executing every node for each run.")
+cache_group.add_argument("--high-ram", action="store_true", help="Can improve performance slightly on high RAM or on systems where pagefile use is preferred over model loading.")
 
 attn_group = parser.add_mutually_exclusive_group()
 attn_group.add_argument("--use-split-cross-attention", action="store_true", help="Use the split cross attention optimization. Ignored when xformers is used.")
@@ -133,7 +134,7 @@ upcast.add_argument("--dont-upcast-attention", action="store_true", help="Disabl
 parser.add_argument("--enable-manager", action="store_true", help="Enable the ComfyUI-Manager feature.")
 manager_group = parser.add_mutually_exclusive_group()
 manager_group.add_argument("--disable-manager-ui", action="store_true", help="Disables only the ComfyUI-Manager UI and endpoints. Scheduled installations and similar background tasks will still operate.")
-manager_group.add_argument("--enable-manager-legacy-ui", action="store_true", help="Enables the legacy UI of ComfyUI-Manager")
+manager_group.add_argument("--enable-manager-legacy-ui", action="store_true", help="Enables the legacy UI of ComfyUI-Manager. Implies --enable-manager.")
 
 
 vram_group = parser.add_mutually_exclusive_group()
@@ -249,6 +250,9 @@ else:
 if args.cache_ram is not None and len(args.cache_ram) > 2:
     parser.error("--cache-ram accepts at most two values: active GB and inactive GB")
 
+if args.high_ram:
+    args.cache_classic = True
+
 if args.windows_standalone_build:
     args.auto_launch = True
 
@@ -258,6 +262,10 @@ if args.disable_auto_launch:
 if args.force_fp16:
     args.fp16_unet = True
 
+# '--enable-manager-legacy-ui' is meaningless unless the manager is enabled, so imply '--enable-manager'.
+if args.enable_manager_legacy_ui:
+    args.enable_manager = True
+
 
 # '--fast' is not provided, use an empty set
 if args.fast is None:
diff --git a/comfy/ldm/ideogram4/model.py b/comfy/ldm/ideogram4/model.py
index b86c65bf0..4ea5b8aaf 100644
--- a/comfy/ldm/ideogram4/model.py
+++ b/comfy/ldm/ideogram4/model.py
@@ -106,11 +106,11 @@ class Ideogram4EmbedScalar(nn.Module):
         self.mlp_in = operations.Linear(dim, dim, bias=True, dtype=dtype, device=device)
         self.mlp_out = operations.Linear(dim, dim, bias=True, dtype=dtype, device=device)
 
-    def forward(self, x):
+    def forward(self, x, dtype):
         x = x.to(torch.float32)
         scaled = 1e4 * (x - self.range_min) / (self.range_max - self.range_min)
         emb = _sinusoidal_embedding(scaled, self.dim)
-        emb = emb.to(self.mlp_in.weight.dtype)
+        emb = emb.to(dtype)
         emb = F.silu(self.mlp_in(emb))
         return self.mlp_out(emb)
 
@@ -161,7 +161,7 @@ class Ideogram4Transformer(nn.Module):
         x = x * output_image_mask
         h = self.input_proj(x) * output_image_mask
 
-        t_cond = self.t_embedding(t)
+        t_cond = self.t_embedding(t, dtype=x.dtype)
         if t.dim() == 1:
             t_cond = t_cond.unsqueeze(1)
         adaln_input = F.silu(self.adaln_proj(t_cond))
diff --git a/comfy/model_management.py b/comfy/model_management.py
index 55ddaab8e..b15d08ba1 100644
--- a/comfy/model_management.py
+++ b/comfy/model_management.py
@@ -643,6 +643,8 @@ def free_pins(size, evict_active=False):
     return freed_total
 
 def ensure_pin_budget(size, evict_active=False):
+    if args.high_ram:
+        return True
     if args.fast_disk:
         shortfall = TOTAL_PINNED_MEMORY + size - MAX_PINNED_MEMORY
     else:
@@ -1496,6 +1498,8 @@ if not args.disable_pinned_memory:
 PINNING_ALLOWED_TYPES = set(["Tensor", "Parameter", "QuantizedTensor"])
 
 def pinned_hostbuf_size(size):
+    if args.high_ram:
+        return max(0, int(size * 2))
     return max(0, int(min(size, MAX_PINNED_MEMORY) * 2))
 
 def discard_cuda_async_error():
diff --git a/comfy/ops.py b/comfy/ops.py
index 3c9912aae..3f088a962 100644
--- a/comfy/ops.py
+++ b/comfy/ops.py
@@ -180,7 +180,7 @@ def cast_modules_with_vbar(comfy_modules, dtype, device, bias_dtype, non_blockin
             if pin is not None:
                 cast_maybe_lowvram_patch([pin], dest, offload_stream)
                 return
-            if signature is None:
+            if signature is None or args.high_ram:
                 comfy.pinned_memory.pin_memory(m, subset=subset, size=size)
                 pin = comfy.pinned_memory.get_pin(m, subset=subset)
             cast_maybe_lowvram_patch(source, pin, offload_stream, xfer_dest2=dest)
diff --git a/comfy_api/latest/_io.py b/comfy_api/latest/_io.py
index 37614a4c3..012fae3ac 100644
--- a/comfy_api/latest/_io.py
+++ b/comfy_api/latest/_io.py
@@ -1400,7 +1400,8 @@ class V3Data(TypedDict):
 class HiddenHolder:
     def __init__(self, unique_id: str, prompt: Any,
                  extra_pnginfo: Any, dynprompt: Any,
-                 auth_token_comfy_org: str, api_key_comfy_org: str, **kwargs):
+                 auth_token_comfy_org: str, api_key_comfy_org: str,
+                 comfy_usage_source: str = None, **kwargs):
         self.unique_id = unique_id
         """UNIQUE_ID is the unique identifier of the node, and matches the id property of the node on the client side. It is commonly used in client-server communications (see messages)."""
         self.prompt = prompt
@@ -1413,6 +1414,8 @@ class HiddenHolder:
         """AUTH_TOKEN_COMFY_ORG is a token acquired from signing into a ComfyOrg account on frontend."""
         self.api_key_comfy_org = api_key_comfy_org
         """API_KEY_COMFY_ORG is an API Key generated by ComfyOrg that allows skipping signing into a ComfyOrg account on frontend."""
+        self.comfy_usage_source = comfy_usage_source
+        """COMFY_USAGE_SOURCE identifies the client that submitted the prompt (e.g. comfyui-frontend, comfy-cli, comfyui-mcp); forwarded to API nodes' upstream requests via the Comfy-Usage-Source header."""
 
     def __getattr__(self, key: str):
         '''If hidden variable not found, return None.'''
@@ -1429,6 +1432,7 @@ class HiddenHolder:
             dynprompt=d.get(Hidden.dynprompt, None),
             auth_token_comfy_org=d.get(Hidden.auth_token_comfy_org, None),
             api_key_comfy_org=d.get(Hidden.api_key_comfy_org, None),
+            comfy_usage_source=d.get(Hidden.comfy_usage_source, None),
         )
 
     @classmethod
@@ -1451,6 +1455,8 @@ class Hidden(str, Enum):
     """AUTH_TOKEN_COMFY_ORG is a token acquired from signing into a ComfyOrg account on frontend."""
     api_key_comfy_org = "API_KEY_COMFY_ORG"
     """API_KEY_COMFY_ORG is an API Key generated by ComfyOrg that allows skipping signing into a ComfyOrg account on frontend."""
+    comfy_usage_source = "COMFY_USAGE_SOURCE"
+    """COMFY_USAGE_SOURCE identifies the client that submitted the prompt (e.g. comfyui-frontend, comfy-cli, comfyui-mcp); forwarded to API nodes' upstream requests via the Comfy-Usage-Source header."""
 
 
 @dataclass
@@ -1654,6 +1660,8 @@ class Schema:
                 self.hidden.append(Hidden.auth_token_comfy_org)
             if Hidden.api_key_comfy_org not in self.hidden:
                 self.hidden.append(Hidden.api_key_comfy_org)
+            if Hidden.comfy_usage_source not in self.hidden:
+                self.hidden.append(Hidden.comfy_usage_source)
         # if is an output_node, will need prompt and extra_pnginfo
         if self.is_output_node:
             if Hidden.prompt not in self.hidden:
diff --git a/comfy_api_nodes/apis/__init__.py b/comfy_api_nodes/apis/__init__.py
index 9c4cfb9b6..9a7049ea2 100644
--- a/comfy_api_nodes/apis/__init__.py
+++ b/comfy_api_nodes/apis/__init__.py
@@ -1310,13 +1310,6 @@ class KlingTaskStatus(str, Enum):
     failed = 'failed'
 
 
-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):
     field_16_9 = '16:9'
     field_9_16 = '9:16'
@@ -5179,7 +5172,7 @@ class KlingText2VideoRequest(BaseModel):
     duration: Optional[KlingVideoGenDuration] = '5'
     external_task_id: Optional[str] = Field(None, description='Customized Task ID')
     mode: Optional[KlingVideoGenMode] = 'std'
-    model_name: Optional[KlingTextToVideoModelName] = 'kling-v1'
+    model_name: Optional[str] = 'kling-v1'
     negative_prompt: Optional[str] = Field(
         None, description='Negative text prompt', max_length=2500
     )
diff --git a/comfy_api_nodes/nodes_bria.py b/comfy_api_nodes/nodes_bria.py
index e138fafa9..090154afb 100644
--- a/comfy_api_nodes/nodes_bria.py
+++ b/comfy_api_nodes/nodes_bria.py
@@ -289,7 +289,7 @@ class BriaRemoveVideoBackground(IO.ComfyNode):
             ],
             is_api_node=True,
             price_badge=IO.PriceBadge(
-                expr="""{"type":"usd","usd":0.14,"format":{"suffix":"/second"}}""",
+                expr="""{"type":"usd","usd":0.0042,"format":{"suffix":"/second"}}""",
             ),
         )
 
@@ -357,7 +357,7 @@ class BriaVideoGreenScreen(IO.ComfyNode):
             ],
             is_api_node=True,
             price_badge=IO.PriceBadge(
-                expr="""{"type":"usd","usd":0.14,"format":{"suffix":"/second"}}""",
+                expr="""{"type":"usd","usd":0.0042,"format":{"suffix":"/second"}}""",
             ),
         )
 
@@ -433,7 +433,7 @@ class BriaVideoReplaceBackground(IO.ComfyNode):
             ],
             is_api_node=True,
             price_badge=IO.PriceBadge(
-                expr="""{"type":"usd","usd":0.14,"format":{"suffix":"/second"}}""",
+                expr="""{"type":"usd","usd":0.0042,"format":{"suffix":"/second"}}""",
             ),
         )
 
@@ -452,7 +452,10 @@ class BriaVideoReplaceBackground(IO.ComfyNode):
             validate_video_duration(background_video, max_duration=60.0)
             background_url = await upload_video_to_comfyapi(cls, background_video, wait_label="Uploading background")
         else:
-            background_url = await upload_image_to_comfyapi(cls, background_image, wait_label="Uploading background")
+            # Bria's replace_background 500s on RGBA, so drop the alpha channel before upload.
+            background_url = await upload_image_to_comfyapi(
+                cls, background_image[:, :, :, :3], wait_label="Uploading background"
+            )
         response = await sync_op(
             cls,
             ApiEndpoint(path="/proxy/bria/v2/video/edit/replace_background", method="POST"),
@@ -530,7 +533,7 @@ class BriaTransparentVideoBackground(IO.ComfyNode):
             ],
             is_api_node=True,
             price_badge=IO.PriceBadge(
-                expr="""{"type":"usd","usd":0.14,"format":{"suffix":"/second"}}""",
+                expr="""{"type":"usd","usd":0.0042,"format":{"suffix":"/second"}}""",
             ),
         )
 
@@ -571,7 +574,7 @@ class BriaExtension(ComfyExtension):
             BriaRemoveImageBackground,
             BriaRemoveVideoBackground,
             BriaVideoGreenScreen,
-            # BriaVideoReplaceBackground,  # server returns Status 500 when we pass background video
+            BriaVideoReplaceBackground,
             BriaTransparentVideoBackground,
         ]
 
diff --git a/comfy_api_nodes/nodes_kling.py b/comfy_api_nodes/nodes_kling.py
index d11e42540..c81d3503d 100644
--- a/comfy_api_nodes/nodes_kling.py
+++ b/comfy_api_nodes/nodes_kling.py
@@ -436,7 +436,7 @@ async def execute_text2video(
             negative_prompt=negative_prompt if negative_prompt else None,
             duration=KlingVideoGenDuration(duration),
             mode=KlingVideoGenMode(model_mode),
-            model_name=KlingVideoGenModelName(model_name),
+            model_name=model_name,
             cfg_scale=cfg_scale,
             aspect_ratio=KlingVideoGenAspectRatio(aspect_ratio),
             camera_control=camera_control,
diff --git a/comfy_api_nodes/nodes_sonilo.py b/comfy_api_nodes/nodes_sonilo.py
index 9ce896ed0..24a9a0b06 100644
--- a/comfy_api_nodes/nodes_sonilo.py
+++ b/comfy_api_nodes/nodes_sonilo.py
@@ -16,7 +16,7 @@ from comfy_api_nodes.util import (
 )
 from comfy_api_nodes.util._helpers import (
     default_base_url,
-    get_auth_header,
+    get_comfy_api_headers,
     get_node_id,
     is_processing_interrupted,
 )
@@ -174,8 +174,7 @@ async def _stream_sonilo_music(
     """POST ``form`` to Sonilo, read the NDJSON stream, and return the first stream's audio bytes."""
     url = urljoin(default_base_url().rstrip("/") + "/", endpoint.path.lstrip("/"))
 
-    headers: dict[str, str] = {}
-    headers.update(get_auth_header(cls))
+    headers = get_comfy_api_headers(cls)
     headers.update(endpoint.headers)
 
     node_id = get_node_id(cls)
diff --git a/comfy_api_nodes/util/_helpers.py b/comfy_api_nodes/util/_helpers.py
index 648defe3d..83cf7b001 100644
--- a/comfy_api_nodes/util/_helpers.py
+++ b/comfy_api_nodes/util/_helpers.py
@@ -9,6 +9,7 @@ from io import BytesIO
 from yarl import URL
 
 from comfy.cli_args import args
+from comfy.deploy_environment import get_deploy_environment
 from comfy.model_management import processing_interrupted
 from comfy_api.latest import IO
 
@@ -35,6 +36,30 @@ def get_auth_header(node_cls: type[IO.ComfyNode]) -> dict[str, str]:
     return {}
 
 
+def get_usage_source(node_cls: type[IO.ComfyNode]) -> str:
+    """Source of the prompt that triggered this API node.
+
+    Defaults to "comfyui-api" when the submitting client didn't identify itself,
+    i.e. a direct API call to this server.
+    """
+    return node_cls.hidden.comfy_usage_source or "comfyui-api"
+
+
+def get_comfy_api_headers(node_cls: type[IO.ComfyNode]) -> dict[str, str]:
+    """Common headers (auth, deploy environment, usage source) for Comfy API requests.
+
+    Centralizes the shared header set so every Comfy API request sends a consistent
+    set and new shared headers only need to be added in one place. Intended for
+    relative/cloud URLs resolved against ``default_base_url()``; because the result
+    includes auth, callers must not attach it to arbitrary absolute/presigned URLs.
+    """
+    return {
+        **get_auth_header(node_cls),
+        "Comfy-Env": get_deploy_environment(),
+        "Comfy-Usage-Source": get_usage_source(node_cls),
+    }
+
+
 def default_base_url() -> str:
     return getattr(args, "comfy_api_base", "https://api.comfy.org")
 
diff --git a/comfy_api_nodes/util/client.py b/comfy_api_nodes/util/client.py
index 57c501724..adcde7bcb 100644
--- a/comfy_api_nodes/util/client.py
+++ b/comfy_api_nodes/util/client.py
@@ -19,12 +19,10 @@ from comfy import utils
 from comfy_api.latest import IO
 from server import PromptServer
 
-from comfy.deploy_environment import get_deploy_environment
-
 from . import request_logger
 from ._helpers import (
     default_base_url,
-    get_auth_header,
+    get_comfy_api_headers,
     get_node_id,
     is_processing_interrupted,
     sleep_with_interrupt,
@@ -645,8 +643,7 @@ async def _request_base(cfg: _RequestConfig, expect_binary: bool):
 
         payload_headers = {"Accept": "*/*"} if expect_binary else {"Accept": "application/json"}
         if not parsed_url.scheme and not parsed_url.netloc:  # is URL relative?
-            payload_headers.update(get_auth_header(cfg.node_cls))
-            payload_headers["Comfy-Env"] = get_deploy_environment()
+            payload_headers.update(get_comfy_api_headers(cfg.node_cls))
         if cfg.endpoint.headers:
             payload_headers.update(cfg.endpoint.headers)
 
diff --git a/comfy_api_nodes/util/download_helpers.py b/comfy_api_nodes/util/download_helpers.py
index aa588d038..0ec3c6e66 100644
--- a/comfy_api_nodes/util/download_helpers.py
+++ b/comfy_api_nodes/util/download_helpers.py
@@ -17,7 +17,7 @@ from folder_paths import get_output_directory
 from . import request_logger
 from ._helpers import (
     default_base_url,
-    get_auth_header,
+    get_comfy_api_headers,
     is_processing_interrupted,
     sleep_with_interrupt,
     to_aiohttp_url,
@@ -64,7 +64,7 @@ async def download_url_to_bytesio(
         if cls is None:
             raise ValueError("For relative 'cloud' paths, the `cls` parameter is required.")
         url = urljoin(default_base_url().rstrip("/") + "/", url.lstrip("/"))
-        headers = get_auth_header(cls)
+        headers = get_comfy_api_headers(cls)
 
     while True:
         attempt += 1
diff --git a/comfy_extras/nodes_flux.py b/comfy_extras/nodes_flux.py
index afc663b22..ef1757ae5 100644
--- a/comfy_extras/nodes_flux.py
+++ b/comfy_extras/nodes_flux.py
@@ -245,6 +245,11 @@ class KV_Attn_Input:
         cache_key = "{}_{}".format(extra_options["block_type"], extra_options["block_index"])
         if cache_key in self.cache:
             kk, vv = self.cache[cache_key]
+
+            # Fix batch size changing.
+            kk = comfy.utils.repeat_to_batch_size(kk, k.shape[0])
+            vv = comfy.utils.repeat_to_batch_size(vv, v.shape[0])
+
             self.set_cache = False
             return {"q": q, "k": torch.cat((k, kk), dim=2), "v": torch.cat((v, vv), dim=2)}
 
diff --git a/execution.py b/execution.py
index e6c6f39d6..9e16e451d 100644
--- a/execution.py
+++ b/execution.py
@@ -200,6 +200,8 @@ def get_input_data(inputs, class_def, unique_id, execution_list=None, dynprompt=
                 hidden_inputs_v3[io.Hidden.auth_token_comfy_org] = extra_data.get("auth_token_comfy_org", None)
             if io.Hidden.api_key_comfy_org.name in hidden:
                 hidden_inputs_v3[io.Hidden.api_key_comfy_org] = extra_data.get("api_key_comfy_org", None)
+            if io.Hidden.comfy_usage_source.name in hidden:
+                hidden_inputs_v3[io.Hidden.comfy_usage_source] = extra_data.get("comfy_usage_source", None)
     else:
         if "hidden" in valid_inputs:
             h = valid_inputs["hidden"]
@@ -216,6 +218,8 @@ def get_input_data(inputs, class_def, unique_id, execution_list=None, dynprompt=
                     input_data_all[x] = [extra_data.get("auth_token_comfy_org", None)]
                 if h[x] == "API_KEY_COMFY_ORG":
                     input_data_all[x] = [extra_data.get("api_key_comfy_org", None)]
+                if h[x] == "COMFY_USAGE_SOURCE":
+                    input_data_all[x] = [extra_data.get("comfy_usage_source", None)]
     v3_data["hidden_inputs"] = hidden_inputs_v3
     return input_data_all, missing_keys, v3_data
 
diff --git a/server.py b/server.py
index cc3b33a5c..ccc92e5ab 100644
--- a/server.py
+++ b/server.py
@@ -971,6 +971,11 @@ class PromptServer():
 
                 if "client_id" in json_data:
                     extra_data["client_id"] = json_data["client_id"]
+
+                if "comfy_usage_source" not in extra_data:
+                    usage_source = request.headers.get("Comfy-Usage-Source")
+                    if usage_source:
+                        extra_data["comfy_usage_source"] = usage_source
                 if valid[0]:
                     outputs_to_execute = valid[2]
                     sensitive = {}