mirror of
https://github.com/comfyanonymous/ComfyUI.git
synced 2026-05-08 19:27:59 +08:00
Compare commits
17 Commits
alexis/cod
...
v0.20.3
| Author | SHA1 | Date | |
|---|---|---|---|
| 491e8454c1 | |||
| c4ace3e25a | |||
| 0a41ce631a | |||
| 81646474d8 | |||
| 9dac8fec7d | |||
| 5c5151cdbe | |||
| d55676e121 | |||
| 43145a9eb7 | |||
| 52baf57d68 | |||
| c40c55ce35 | |||
| 5a252d437e | |||
| 0aaa202a1d | |||
| e83cf74957 | |||
| 12ff560e7d | |||
| de1c50e214 | |||
| 7f4c0188e9 | |||
| f3f9615a80 |
@ -342,6 +342,12 @@ def model_lora_keys_unet(model, key_map={}):
|
||||
key_map["base_model.model.{}".format(key_lora)] = k # Official base model loras
|
||||
key_map["lycoris_{}".format(key_lora.replace(".", "_"))] = k # LyCORIS/LoKR format
|
||||
|
||||
if isinstance(model, comfy.model_base.ErnieImage):
|
||||
for k in sdk:
|
||||
if k.startswith("diffusion_model.") and k.endswith(".weight"):
|
||||
key_lora = k[len("diffusion_model."):-len(".weight")]
|
||||
key_map["transformer.{}".format(key_lora)] = k
|
||||
|
||||
return key_map
|
||||
|
||||
|
||||
|
||||
@ -157,6 +157,11 @@ class SeedanceCreateAssetResponse(BaseModel):
|
||||
asset_id: str = Field(...)
|
||||
|
||||
|
||||
class SeedanceVirtualLibraryCreateAssetRequest(BaseModel):
|
||||
url: str = Field(..., description="Publicly accessible URL of the image asset to upload.")
|
||||
hash: str = Field(..., description="Dedup key. Re-submitting the same hash returns the existing asset id.")
|
||||
|
||||
|
||||
# Dollars per 1K tokens, keyed by (model_id, has_video_input).
|
||||
SEEDANCE2_PRICE_PER_1K_TOKENS = {
|
||||
("dreamina-seedance-2-0-260128", False): 0.007,
|
||||
|
||||
@ -1,15 +1,12 @@
|
||||
from __future__ import annotations
|
||||
|
||||
|
||||
import torch
|
||||
|
||||
from enum import Enum
|
||||
from typing import Optional, Union
|
||||
|
||||
import torch
|
||||
from pydantic import BaseModel, Field, confloat
|
||||
|
||||
|
||||
|
||||
class LumaIO:
|
||||
LUMA_REF = "LUMA_REF"
|
||||
LUMA_CONCEPTS = "LUMA_CONCEPTS"
|
||||
@ -183,13 +180,13 @@ class LumaAssets(BaseModel):
|
||||
|
||||
|
||||
class LumaImageRef(BaseModel):
|
||||
'''Used for image gen'''
|
||||
"""Used for image gen"""
|
||||
url: str = Field(..., description='The URL of the image reference')
|
||||
weight: confloat(ge=0.0, le=1.0) = Field(..., description='The weight of the image reference')
|
||||
|
||||
|
||||
class LumaImageReference(BaseModel):
|
||||
'''Used for video gen'''
|
||||
"""Used for video gen"""
|
||||
type: Optional[str] = Field('image', description='Input type, defaults to image')
|
||||
url: str = Field(..., description='The URL of the image')
|
||||
|
||||
@ -251,3 +248,32 @@ class LumaGeneration(BaseModel):
|
||||
assets: Optional[LumaAssets] = Field(None, description='The assets of the generation')
|
||||
model: str = Field(..., description='The model used for the generation')
|
||||
request: Union[LumaGenerationRequest, LumaImageGenerationRequest] = Field(..., description="The request used for the generation")
|
||||
|
||||
|
||||
class Luma2ImageRef(BaseModel):
|
||||
url: str | None = None
|
||||
data: str | None = None
|
||||
media_type: str | None = None
|
||||
|
||||
|
||||
class Luma2GenerationRequest(BaseModel):
|
||||
prompt: str = Field(..., min_length=1, max_length=6000)
|
||||
model: str | None = None
|
||||
type: str | None = None
|
||||
aspect_ratio: str | None = None
|
||||
style: str | None = None
|
||||
output_format: str | None = None
|
||||
web_search: bool | None = None
|
||||
image_ref: list[Luma2ImageRef] | None = None
|
||||
source: Luma2ImageRef | None = None
|
||||
|
||||
|
||||
class Luma2Generation(BaseModel):
|
||||
id: str | None = None
|
||||
type: str | None = None
|
||||
state: str | None = None
|
||||
model: str | None = None
|
||||
created_at: str | None = None
|
||||
output: list[LumaImageReference] | None = None
|
||||
failure_reason: str | None = None
|
||||
failure_code: str | None = None
|
||||
|
||||
@ -1,152 +0,0 @@
|
||||
from enum import Enum
|
||||
from typing import Optional, Dict, Any
|
||||
|
||||
from pydantic import BaseModel, Field, StrictBytes
|
||||
|
||||
|
||||
class MoonvalleyPromptResponse(BaseModel):
|
||||
error: Optional[Dict[str, Any]] = None
|
||||
frame_conditioning: Optional[Dict[str, Any]] = None
|
||||
id: Optional[str] = None
|
||||
inference_params: Optional[Dict[str, Any]] = None
|
||||
meta: Optional[Dict[str, Any]] = None
|
||||
model_params: Optional[Dict[str, Any]] = None
|
||||
output_url: Optional[str] = None
|
||||
prompt_text: Optional[str] = None
|
||||
status: Optional[str] = None
|
||||
|
||||
|
||||
class MoonvalleyTextToVideoInferenceParams(BaseModel):
|
||||
add_quality_guidance: Optional[bool] = Field(
|
||||
True, description='Whether to add quality guidance'
|
||||
)
|
||||
caching_coefficient: Optional[float] = Field(
|
||||
0.3, description='Caching coefficient for optimization'
|
||||
)
|
||||
caching_cooldown: Optional[int] = Field(
|
||||
3, description='Number of caching cooldown steps'
|
||||
)
|
||||
caching_warmup: Optional[int] = Field(
|
||||
3, description='Number of caching warmup steps'
|
||||
)
|
||||
clip_value: Optional[float] = Field(
|
||||
3, description='CLIP value for generation control'
|
||||
)
|
||||
conditioning_frame_index: Optional[int] = Field(
|
||||
0, description='Index of the conditioning frame'
|
||||
)
|
||||
cooldown_steps: Optional[int] = Field(
|
||||
75, description='Number of cooldown steps (calculated based on num_frames)'
|
||||
)
|
||||
fps: Optional[int] = Field(
|
||||
24, description='Frames per second of the generated video'
|
||||
)
|
||||
guidance_scale: Optional[float] = Field(
|
||||
10, description='Guidance scale for generation control'
|
||||
)
|
||||
height: Optional[int] = Field(
|
||||
1080, description='Height of the generated video in pixels'
|
||||
)
|
||||
negative_prompt: Optional[str] = Field(None, description='Negative prompt text')
|
||||
num_frames: Optional[int] = Field(64, description='Number of frames to generate')
|
||||
seed: Optional[int] = Field(
|
||||
None, description='Random seed for generation (default: random)'
|
||||
)
|
||||
shift_value: Optional[float] = Field(
|
||||
3, description='Shift value for generation control'
|
||||
)
|
||||
steps: Optional[int] = Field(80, description='Number of denoising steps')
|
||||
use_guidance_schedule: Optional[bool] = Field(
|
||||
True, description='Whether to use guidance scheduling'
|
||||
)
|
||||
use_negative_prompts: Optional[bool] = Field(
|
||||
False, description='Whether to use negative prompts'
|
||||
)
|
||||
use_timestep_transform: Optional[bool] = Field(
|
||||
True, description='Whether to use timestep transformation'
|
||||
)
|
||||
warmup_steps: Optional[int] = Field(
|
||||
0, description='Number of warmup steps (calculated based on num_frames)'
|
||||
)
|
||||
width: Optional[int] = Field(
|
||||
1920, description='Width of the generated video in pixels'
|
||||
)
|
||||
|
||||
|
||||
class MoonvalleyTextToVideoRequest(BaseModel):
|
||||
image_url: Optional[str] = None
|
||||
inference_params: Optional[MoonvalleyTextToVideoInferenceParams] = None
|
||||
prompt_text: Optional[str] = None
|
||||
webhook_url: Optional[str] = None
|
||||
|
||||
|
||||
class MoonvalleyUploadFileRequest(BaseModel):
|
||||
file: Optional[StrictBytes] = None
|
||||
|
||||
|
||||
class MoonvalleyUploadFileResponse(BaseModel):
|
||||
access_url: Optional[str] = None
|
||||
|
||||
|
||||
class MoonvalleyVideoToVideoInferenceParams(BaseModel):
|
||||
add_quality_guidance: Optional[bool] = Field(
|
||||
True, description='Whether to add quality guidance'
|
||||
)
|
||||
caching_coefficient: Optional[float] = Field(
|
||||
0.3, description='Caching coefficient for optimization'
|
||||
)
|
||||
caching_cooldown: Optional[int] = Field(
|
||||
3, description='Number of caching cooldown steps'
|
||||
)
|
||||
caching_warmup: Optional[int] = Field(
|
||||
3, description='Number of caching warmup steps'
|
||||
)
|
||||
clip_value: Optional[float] = Field(
|
||||
3, description='CLIP value for generation control'
|
||||
)
|
||||
conditioning_frame_index: Optional[int] = Field(
|
||||
0, description='Index of the conditioning frame'
|
||||
)
|
||||
cooldown_steps: Optional[int] = Field(
|
||||
36, description='Number of cooldown steps (calculated based on num_frames)'
|
||||
)
|
||||
guidance_scale: Optional[float] = Field(
|
||||
15, description='Guidance scale for generation control'
|
||||
)
|
||||
negative_prompt: Optional[str] = Field(None, description='Negative prompt text')
|
||||
seed: Optional[int] = Field(
|
||||
None, description='Random seed for generation (default: random)'
|
||||
)
|
||||
shift_value: Optional[float] = Field(
|
||||
3, description='Shift value for generation control'
|
||||
)
|
||||
steps: Optional[int] = Field(80, description='Number of denoising steps')
|
||||
use_guidance_schedule: Optional[bool] = Field(
|
||||
True, description='Whether to use guidance scheduling'
|
||||
)
|
||||
use_negative_prompts: Optional[bool] = Field(
|
||||
False, description='Whether to use negative prompts'
|
||||
)
|
||||
use_timestep_transform: Optional[bool] = Field(
|
||||
True, description='Whether to use timestep transformation'
|
||||
)
|
||||
warmup_steps: Optional[int] = Field(
|
||||
24, description='Number of warmup steps (calculated based on num_frames)'
|
||||
)
|
||||
|
||||
|
||||
class ControlType(str, Enum):
|
||||
motion_control = 'motion_control'
|
||||
pose_control = 'pose_control'
|
||||
|
||||
|
||||
class MoonvalleyVideoToVideoRequest(BaseModel):
|
||||
control_type: ControlType = Field(
|
||||
..., description='Supported types for video control'
|
||||
)
|
||||
inference_params: Optional[MoonvalleyVideoToVideoInferenceParams] = None
|
||||
prompt_text: str = Field(..., description='Describes the video to generate')
|
||||
video_url: str = Field(..., description='Url to control video')
|
||||
webhook_url: Optional[str] = Field(
|
||||
None, description='Optional webhook URL for notifications'
|
||||
)
|
||||
@ -56,14 +56,14 @@ class ModelResponseProperties(BaseModel):
|
||||
instructions: str | None = Field(None)
|
||||
max_output_tokens: int | None = Field(None)
|
||||
model: str | None = Field(None)
|
||||
temperature: float | None = Field(1, description="Controls randomness in the response", ge=0.0, le=2.0)
|
||||
temperature: float | None = Field(None, description="Controls randomness in the response", ge=0.0, le=2.0)
|
||||
top_p: float | None = Field(
|
||||
1,
|
||||
None,
|
||||
description="Controls diversity of the response via nucleus sampling",
|
||||
ge=0.0,
|
||||
le=1.0,
|
||||
)
|
||||
truncation: str | None = Field("disabled", description="Allowed values: 'auto' or 'disabled'")
|
||||
truncation: str | None = Field(None, description="Allowed values: 'auto' or 'disabled'")
|
||||
|
||||
|
||||
class ResponseProperties(BaseModel):
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
from typing import Optional, Union
|
||||
from typing import Optional
|
||||
|
||||
from pydantic import BaseModel, Field
|
||||
|
||||
@ -72,8 +72,11 @@ class VideoEnhancementFilter(BaseModel):
|
||||
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")
|
||||
creativity: float | str | None = Field(None, description="slc-1/slp-2.5: enum (low/middle/high). ast-2: decimal 0.0-1.0.")
|
||||
isOptimizedMode: Optional[bool] = Field(None, description="Set to true for Starlight Creative (slc-1) only")
|
||||
prompt: str | None = Field(None, description="Descriptive scene prompt (ast-2 only)")
|
||||
sharp: float | None = Field(None, description="ast-2 pre-enhance sharpness")
|
||||
realism: float | None = Field(None, description="ast-2 realism control")
|
||||
|
||||
|
||||
class OutputInformationVideo(BaseModel):
|
||||
@ -90,7 +93,7 @@ class Overrides(BaseModel):
|
||||
|
||||
class CreateVideoRequest(BaseModel):
|
||||
source: CreateVideoRequestSource = Field(...)
|
||||
filters: list[Union[VideoFrameInterpolationFilter, VideoEnhancementFilter]] = Field(...)
|
||||
filters: list[VideoFrameInterpolationFilter | VideoEnhancementFilter] = Field(...)
|
||||
output: OutputInformationVideo = Field(...)
|
||||
overrides: Overrides = Field(Overrides(isPaidDiffusion=True))
|
||||
|
||||
|
||||
@ -1,3 +1,4 @@
|
||||
import hashlib
|
||||
import logging
|
||||
import math
|
||||
import re
|
||||
@ -20,6 +21,7 @@ from comfy_api_nodes.apis.bytedance import (
|
||||
SeedanceCreateAssetResponse,
|
||||
SeedanceCreateVisualValidateSessionResponse,
|
||||
SeedanceGetVisualValidateSessionResponse,
|
||||
SeedanceVirtualLibraryCreateAssetRequest,
|
||||
Seedream4Options,
|
||||
Seedream4TaskCreationRequest,
|
||||
TaskAudioContent,
|
||||
@ -271,6 +273,30 @@ async def _wait_for_asset_active(cls: type[IO.ComfyNode], asset_id: str, group_i
|
||||
)
|
||||
|
||||
|
||||
async def _seedance_virtual_library_upload_image_asset(
|
||||
cls: type[IO.ComfyNode],
|
||||
image: torch.Tensor,
|
||||
*,
|
||||
wait_label: str = "Uploading image",
|
||||
) -> str:
|
||||
"""Upload an image into the caller's per-customer Seedance virtual library."""
|
||||
public_url = await upload_image_to_comfyapi(cls, image, wait_label=wait_label)
|
||||
normalized = image.detach().cpu().contiguous().to(torch.float32)
|
||||
digest = hashlib.sha256()
|
||||
digest.update(str(tuple(normalized.shape)).encode("utf-8"))
|
||||
digest.update(b"\0")
|
||||
digest.update(normalized.numpy().tobytes())
|
||||
image_hash = digest.hexdigest()
|
||||
create_resp = await sync_op(
|
||||
cls,
|
||||
ApiEndpoint(path="/proxy/seedance/virtual-library/assets", method="POST"),
|
||||
response_model=SeedanceCreateAssetResponse,
|
||||
data=SeedanceVirtualLibraryCreateAssetRequest(url=public_url, hash=image_hash),
|
||||
)
|
||||
await _wait_for_asset_active(cls, create_resp.asset_id, group_id="virtual-library")
|
||||
return f"asset://{create_resp.asset_id}"
|
||||
|
||||
|
||||
def _seedance2_price_extractor(model_id: str, has_video_input: bool):
|
||||
"""Returns a price_extractor closure for Seedance 2.0 poll_op."""
|
||||
rate = SEEDANCE2_PRICE_PER_1K_TOKENS.get((model_id, has_video_input))
|
||||
@ -1377,7 +1403,6 @@ class ByteDance2TextToVideoNode(IO.ComfyNode):
|
||||
status_extractor=lambda r: r.status,
|
||||
price_extractor=_seedance2_price_extractor(model_id, has_video_input=False),
|
||||
poll_interval=9,
|
||||
max_poll_attempts=180,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(response.content.video_url))
|
||||
|
||||
@ -1507,7 +1532,9 @@ class ByteDance2FirstLastFrameNode(IO.ComfyNode):
|
||||
if first_frame_asset_id:
|
||||
first_frame_url = image_assets[first_frame_asset_id]
|
||||
else:
|
||||
first_frame_url = await upload_image_to_comfyapi(cls, first_frame, wait_label="Uploading first frame.")
|
||||
first_frame_url = await _seedance_virtual_library_upload_image_asset(
|
||||
cls, first_frame, wait_label="Uploading first frame."
|
||||
)
|
||||
|
||||
content: list[TaskTextContent | TaskImageContent] = [
|
||||
TaskTextContent(text=model["prompt"]),
|
||||
@ -1527,7 +1554,9 @@ class ByteDance2FirstLastFrameNode(IO.ComfyNode):
|
||||
content.append(
|
||||
TaskImageContent(
|
||||
image_url=TaskImageContentUrl(
|
||||
url=await upload_image_to_comfyapi(cls, last_frame, wait_label="Uploading last frame.")
|
||||
url=await _seedance_virtual_library_upload_image_asset(
|
||||
cls, last_frame, wait_label="Uploading last frame."
|
||||
)
|
||||
),
|
||||
role="last_frame",
|
||||
),
|
||||
@ -1555,7 +1584,6 @@ class ByteDance2FirstLastFrameNode(IO.ComfyNode):
|
||||
status_extractor=lambda r: r.status,
|
||||
price_extractor=_seedance2_price_extractor(model_id, has_video_input=False),
|
||||
poll_interval=9,
|
||||
max_poll_attempts=180,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(response.content.video_url))
|
||||
|
||||
@ -1805,9 +1833,9 @@ class ByteDance2ReferenceNode(IO.ComfyNode):
|
||||
content.append(
|
||||
TaskImageContent(
|
||||
image_url=TaskImageContentUrl(
|
||||
url=await upload_image_to_comfyapi(
|
||||
url=await _seedance_virtual_library_upload_image_asset(
|
||||
cls,
|
||||
image=reference_images[key],
|
||||
reference_images[key],
|
||||
wait_label=f"Uploading image {i}",
|
||||
),
|
||||
),
|
||||
@ -1877,7 +1905,6 @@ class ByteDance2ReferenceNode(IO.ComfyNode):
|
||||
status_extractor=lambda r: r.status,
|
||||
price_extractor=_seedance2_price_extractor(model_id, has_video_input=has_video_input),
|
||||
poll_interval=9,
|
||||
max_poll_attempts=180,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(response.content.video_url))
|
||||
|
||||
|
||||
@ -83,13 +83,16 @@ class GeminiImageModel(str, Enum):
|
||||
|
||||
async def create_image_parts(
|
||||
cls: type[IO.ComfyNode],
|
||||
images: Input.Image,
|
||||
images: Input.Image | list[Input.Image],
|
||||
image_limit: int = 0,
|
||||
) -> list[GeminiPart]:
|
||||
image_parts: list[GeminiPart] = []
|
||||
if image_limit < 0:
|
||||
raise ValueError("image_limit must be greater than or equal to 0 when creating Gemini image parts.")
|
||||
total_images = get_number_of_images(images)
|
||||
|
||||
# Accept either a single (possibly-batched) tensor or a list of them; share URL budget across all.
|
||||
images_list: list[Input.Image] = images if isinstance(images, list) else [images]
|
||||
total_images = sum(get_number_of_images(img) for img in images_list)
|
||||
if total_images <= 0:
|
||||
raise ValueError("No images provided to create_image_parts; at least one image is required.")
|
||||
|
||||
@ -98,10 +101,18 @@ async def create_image_parts(
|
||||
|
||||
# Number of images we'll send as URLs (fileData)
|
||||
num_url_images = min(effective_max, 10) # Vertex API max number of image links
|
||||
upload_kwargs: dict = {"wait_label": "Uploading reference images"}
|
||||
if effective_max > num_url_images:
|
||||
# Split path (e.g. 11+ images): suppress per-image counter to avoid a confusing dual-fraction label.
|
||||
upload_kwargs = {
|
||||
"wait_label": f"Uploading reference images ({num_url_images}+)",
|
||||
"show_batch_index": False,
|
||||
}
|
||||
reference_images_urls = await upload_images_to_comfyapi(
|
||||
cls,
|
||||
images,
|
||||
images_list,
|
||||
max_images=num_url_images,
|
||||
**upload_kwargs,
|
||||
)
|
||||
for reference_image_url in reference_images_urls:
|
||||
image_parts.append(
|
||||
@ -112,15 +123,22 @@ async def create_image_parts(
|
||||
)
|
||||
)
|
||||
)
|
||||
for idx in range(num_url_images, effective_max):
|
||||
image_parts.append(
|
||||
GeminiPart(
|
||||
inlineData=GeminiInlineData(
|
||||
mimeType=GeminiMimeType.image_png,
|
||||
data=tensor_to_base64_string(images[idx]),
|
||||
if effective_max > num_url_images:
|
||||
flat: list[torch.Tensor] = []
|
||||
for tensor in images_list:
|
||||
if len(tensor.shape) == 4:
|
||||
flat.extend(tensor[i] for i in range(tensor.shape[0]))
|
||||
else:
|
||||
flat.append(tensor)
|
||||
for idx in range(num_url_images, effective_max):
|
||||
image_parts.append(
|
||||
GeminiPart(
|
||||
inlineData=GeminiInlineData(
|
||||
mimeType=GeminiMimeType.image_png,
|
||||
data=tensor_to_base64_string(flat[idx]),
|
||||
)
|
||||
)
|
||||
)
|
||||
)
|
||||
return image_parts
|
||||
|
||||
|
||||
@ -891,10 +909,6 @@ class GeminiNanoBanana2(IO.ComfyNode):
|
||||
"9:16",
|
||||
"16:9",
|
||||
"21:9",
|
||||
# "1:4",
|
||||
# "4:1",
|
||||
# "8:1",
|
||||
# "1:8",
|
||||
],
|
||||
default="auto",
|
||||
tooltip="If set to 'auto', matches your input image's aspect ratio; "
|
||||
@ -902,12 +916,7 @@ class GeminiNanoBanana2(IO.ComfyNode):
|
||||
),
|
||||
IO.Combo.Input(
|
||||
"resolution",
|
||||
options=[
|
||||
# "512px",
|
||||
"1K",
|
||||
"2K",
|
||||
"4K",
|
||||
],
|
||||
options=["1K", "2K", "4K"],
|
||||
tooltip="Target output resolution. For 2K/4K the native Gemini upscaler is used.",
|
||||
),
|
||||
IO.Combo.Input(
|
||||
@ -956,6 +965,7 @@ class GeminiNanoBanana2(IO.ComfyNode):
|
||||
],
|
||||
is_api_node=True,
|
||||
price_badge=GEMINI_IMAGE_2_PRICE_BADGE,
|
||||
is_deprecated=True,
|
||||
)
|
||||
|
||||
@classmethod
|
||||
@ -1016,6 +1026,197 @@ class GeminiNanoBanana2(IO.ComfyNode):
|
||||
)
|
||||
|
||||
|
||||
def _nano_banana_2_v2_model_inputs():
|
||||
return [
|
||||
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",
|
||||
"1:4",
|
||||
"4:1",
|
||||
"8:1",
|
||||
"1:8",
|
||||
],
|
||||
default="auto",
|
||||
tooltip="If set to 'auto', matches your input image's aspect ratio; "
|
||||
"if no image is provided, a 16:9 square is usually generated.",
|
||||
),
|
||||
IO.Combo.Input(
|
||||
"resolution",
|
||||
options=["1K", "2K", "4K"],
|
||||
tooltip="Target output resolution. For 2K/4K the native Gemini upscaler is used.",
|
||||
),
|
||||
IO.Combo.Input(
|
||||
"thinking_level",
|
||||
options=["MINIMAL", "HIGH"],
|
||||
),
|
||||
IO.Autogrow.Input(
|
||||
"images",
|
||||
template=IO.Autogrow.TemplateNames(
|
||||
IO.Image.Input("image"),
|
||||
names=[f"image_{i}" for i in range(1, 15)],
|
||||
min=0,
|
||||
),
|
||||
tooltip="Optional reference image(s). Up to 14 images total.",
|
||||
),
|
||||
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.",
|
||||
),
|
||||
]
|
||||
|
||||
|
||||
class GeminiNanoBanana2V2(IO.ComfyNode):
|
||||
|
||||
@classmethod
|
||||
def define_schema(cls):
|
||||
return IO.Schema(
|
||||
node_id="GeminiNanoBanana2V2",
|
||||
display_name="Nano Banana 2",
|
||||
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.DynamicCombo.Input(
|
||||
"model",
|
||||
options=[
|
||||
IO.DynamicCombo.Option(
|
||||
"Nano Banana 2 (Gemini 3.1 Flash Image)",
|
||||
_nano_banana_2_v2_model_inputs(),
|
||||
),
|
||||
],
|
||||
),
|
||||
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(
|
||||
"response_modalities",
|
||||
options=["IMAGE", "IMAGE+TEXT"],
|
||||
advanced=True,
|
||||
),
|
||||
IO.String.Input(
|
||||
"system_prompt",
|
||||
multiline=True,
|
||||
default=GEMINI_IMAGE_SYS_PROMPT,
|
||||
optional=True,
|
||||
tooltip="Foundational instructions that dictate an AI's behavior.",
|
||||
advanced=True,
|
||||
),
|
||||
],
|
||||
outputs=[
|
||||
IO.Image.Output(),
|
||||
IO.String.Output(),
|
||||
IO.Image.Output(
|
||||
display_name="thought_image",
|
||||
tooltip="First image from the model's thinking process. "
|
||||
"Only available with thinking_level HIGH and IMAGE+TEXT modality.",
|
||||
),
|
||||
],
|
||||
hidden=[
|
||||
IO.Hidden.auth_token_comfy_org,
|
||||
IO.Hidden.api_key_comfy_org,
|
||||
IO.Hidden.unique_id,
|
||||
],
|
||||
is_api_node=True,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["model", "model.resolution"]),
|
||||
expr="""
|
||||
(
|
||||
$r := $lookup(widgets, "model.resolution");
|
||||
$prices := {"1k": 0.0696, "2k": 0.1014, "4k": 0.154};
|
||||
{"type":"usd","usd": $lookup($prices, $r), "format":{"suffix":"/Image","approximate":true}}
|
||||
)
|
||||
""",
|
||||
),
|
||||
)
|
||||
|
||||
@classmethod
|
||||
async def execute(
|
||||
cls,
|
||||
prompt: str,
|
||||
model: dict,
|
||||
seed: int,
|
||||
response_modalities: str,
|
||||
system_prompt: str = "",
|
||||
) -> IO.NodeOutput:
|
||||
validate_string(prompt, strip_whitespace=True, min_length=1)
|
||||
model_choice = model["model"]
|
||||
if model_choice == "Nano Banana 2 (Gemini 3.1 Flash Image)":
|
||||
model_id = "gemini-3.1-flash-image-preview"
|
||||
else:
|
||||
model_id = model_choice
|
||||
|
||||
images = model.get("images") or {}
|
||||
parts: list[GeminiPart] = [GeminiPart(text=prompt)]
|
||||
if images:
|
||||
image_tensors: list[Input.Image] = [t for t in images.values() if t is not None]
|
||||
if image_tensors:
|
||||
if sum(get_number_of_images(t) for t in image_tensors) > 14:
|
||||
raise ValueError("The current maximum number of supported images is 14.")
|
||||
parts.extend(await create_image_parts(cls, image_tensors))
|
||||
files = model.get("files")
|
||||
if files is not None:
|
||||
parts.extend(files)
|
||||
|
||||
image_config = GeminiImageConfig(imageSize=model["resolution"])
|
||||
if model["aspect_ratio"] != "auto":
|
||||
image_config.aspectRatio = model["aspect_ratio"]
|
||||
|
||||
gemini_system_prompt = None
|
||||
if system_prompt:
|
||||
gemini_system_prompt = GeminiSystemInstructionContent(parts=[GeminiTextPart(text=system_prompt)], role=None)
|
||||
|
||||
response = await sync_op(
|
||||
cls,
|
||||
ApiEndpoint(path=f"/proxy/vertexai/gemini/{model_id}", 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,
|
||||
thinkingConfig=GeminiThinkingConfig(thinkingLevel=model["thinking_level"]),
|
||||
),
|
||||
systemInstruction=gemini_system_prompt,
|
||||
),
|
||||
response_model=GeminiGenerateContentResponse,
|
||||
price_extractor=calculate_tokens_price,
|
||||
)
|
||||
return IO.NodeOutput(
|
||||
await get_image_from_response(response),
|
||||
get_text_from_response(response),
|
||||
await get_image_from_response(response, thought=True),
|
||||
)
|
||||
|
||||
|
||||
class GeminiExtension(ComfyExtension):
|
||||
@override
|
||||
async def get_node_list(self) -> list[type[IO.ComfyNode]]:
|
||||
@ -1024,6 +1225,7 @@ class GeminiExtension(ComfyExtension):
|
||||
GeminiImage,
|
||||
GeminiImage2,
|
||||
GeminiNanoBanana2,
|
||||
GeminiNanoBanana2V2,
|
||||
GeminiInputFiles,
|
||||
]
|
||||
|
||||
|
||||
@ -54,7 +54,12 @@ class GrokImageNode(IO.ComfyNode):
|
||||
inputs=[
|
||||
IO.Combo.Input(
|
||||
"model",
|
||||
options=["grok-imagine-image-pro", "grok-imagine-image", "grok-imagine-image-beta"],
|
||||
options=[
|
||||
"grok-imagine-image-quality",
|
||||
"grok-imagine-image-pro",
|
||||
"grok-imagine-image",
|
||||
"grok-imagine-image-beta",
|
||||
],
|
||||
),
|
||||
IO.String.Input(
|
||||
"prompt",
|
||||
@ -111,10 +116,12 @@ class GrokImageNode(IO.ComfyNode):
|
||||
],
|
||||
is_api_node=True,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["model", "number_of_images"]),
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["model", "number_of_images", "resolution"]),
|
||||
expr="""
|
||||
(
|
||||
$rate := $contains(widgets.model, "pro") ? 0.07 : 0.02;
|
||||
$rate := widgets.model = "grok-imagine-image-quality"
|
||||
? (widgets.resolution = "1k" ? 0.05 : 0.07)
|
||||
: ($contains(widgets.model, "pro") ? 0.07 : 0.02);
|
||||
{"type":"usd","usd": $rate * widgets.number_of_images}
|
||||
)
|
||||
""",
|
||||
@ -167,7 +174,12 @@ class GrokImageEditNode(IO.ComfyNode):
|
||||
inputs=[
|
||||
IO.Combo.Input(
|
||||
"model",
|
||||
options=["grok-imagine-image-pro", "grok-imagine-image", "grok-imagine-image-beta"],
|
||||
options=[
|
||||
"grok-imagine-image-quality",
|
||||
"grok-imagine-image-pro",
|
||||
"grok-imagine-image",
|
||||
"grok-imagine-image-beta",
|
||||
],
|
||||
),
|
||||
IO.Image.Input("image", display_name="images"),
|
||||
IO.String.Input(
|
||||
@ -228,11 +240,19 @@ class GrokImageEditNode(IO.ComfyNode):
|
||||
],
|
||||
is_api_node=True,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["model", "number_of_images"]),
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["model", "number_of_images", "resolution"]),
|
||||
expr="""
|
||||
(
|
||||
$rate := $contains(widgets.model, "pro") ? 0.07 : 0.02;
|
||||
{"type":"usd","usd": 0.002 + $rate * widgets.number_of_images}
|
||||
$isQualityModel := widgets.model = "grok-imagine-image-quality";
|
||||
$isPro := $contains(widgets.model, "pro");
|
||||
$rate := $isQualityModel
|
||||
? (widgets.resolution = "1k" ? 0.05 : 0.07)
|
||||
: ($isPro ? 0.07 : 0.02);
|
||||
$base := $isQualityModel ? 0.01 : 0.002;
|
||||
$output := $rate * widgets.number_of_images;
|
||||
$isPro
|
||||
? {"type":"usd","usd": $base + $output}
|
||||
: {"type":"range_usd","min_usd": $base + $output, "max_usd": 3 * $base + $output}
|
||||
)
|
||||
""",
|
||||
),
|
||||
|
||||
@ -178,7 +178,6 @@ class HitPawGeneralImageEnhance(IO.ComfyNode):
|
||||
status_extractor=lambda x: x.data.status,
|
||||
price_extractor=lambda x: request_price,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_image_tensor(final_response.data.res_url))
|
||||
|
||||
@ -324,7 +323,6 @@ class HitPawVideoEnhance(IO.ComfyNode):
|
||||
status_extractor=lambda x: x.data.status,
|
||||
price_extractor=lambda x: request_price,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=320,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(final_response.data.res_url))
|
||||
|
||||
|
||||
@ -276,7 +276,6 @@ async def finish_omni_video_task(cls: type[IO.ComfyNode], response: TaskStatusRe
|
||||
cls,
|
||||
ApiEndpoint(path=f"/proxy/kling/v1/videos/omni-video/{response.data.task_id}"),
|
||||
response_model=TaskStatusResponse,
|
||||
max_poll_attempts=280,
|
||||
status_extractor=lambda r: (r.data.task_status if r.data else None),
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(final_response.data.task_result.videos[0].url))
|
||||
@ -2788,11 +2787,15 @@ class MotionControl(IO.ComfyNode):
|
||||
],
|
||||
is_api_node=True,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["mode"]),
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["mode", "model"]),
|
||||
expr="""
|
||||
(
|
||||
$prices := {"std": 0.07, "pro": 0.112};
|
||||
{"type":"usd","usd": $lookup($prices, widgets.mode), "format":{"suffix":"/second"}}
|
||||
$prices := {
|
||||
"kling-v3": {"std": 0.126, "pro": 0.168},
|
||||
"kling-v2-6": {"std": 0.07, "pro": 0.112}
|
||||
};
|
||||
$modelPrices := $lookup($prices, widgets.model);
|
||||
{"type":"usd","usd": $lookup($modelPrices, widgets.mode), "format":{"suffix":"/second"}}
|
||||
)
|
||||
""",
|
||||
),
|
||||
@ -3062,7 +3065,6 @@ class KlingVideoNode(IO.ComfyNode):
|
||||
cls,
|
||||
ApiEndpoint(path=poll_path),
|
||||
response_model=TaskStatusResponse,
|
||||
max_poll_attempts=280,
|
||||
status_extractor=lambda r: (r.data.task_status if r.data else None),
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(final_response.data.task_result.videos[0].url))
|
||||
@ -3188,7 +3190,6 @@ class KlingFirstLastFrameNode(IO.ComfyNode):
|
||||
cls,
|
||||
ApiEndpoint(path=f"/proxy/kling/v1/videos/image2video/{response.data.task_id}"),
|
||||
response_model=TaskStatusResponse,
|
||||
max_poll_attempts=280,
|
||||
status_extractor=lambda r: (r.data.task_status if r.data else None),
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(final_response.data.task_result.videos[0].url))
|
||||
|
||||
@ -1,10 +1,11 @@
|
||||
from typing import Optional
|
||||
|
||||
import torch
|
||||
from typing_extensions import override
|
||||
|
||||
from comfy_api.latest import IO, ComfyExtension
|
||||
from comfy_api.latest import IO, ComfyExtension, Input
|
||||
from comfy_api_nodes.apis.luma import (
|
||||
Luma2Generation,
|
||||
Luma2GenerationRequest,
|
||||
Luma2ImageRef,
|
||||
LumaAspectRatio,
|
||||
LumaCharacterRef,
|
||||
LumaConceptChain,
|
||||
@ -30,6 +31,7 @@ from comfy_api_nodes.util import (
|
||||
download_url_to_video_output,
|
||||
poll_op,
|
||||
sync_op,
|
||||
upload_image_to_comfyapi,
|
||||
upload_images_to_comfyapi,
|
||||
validate_string,
|
||||
)
|
||||
@ -212,9 +214,9 @@ class LumaImageGenerationNode(IO.ComfyNode):
|
||||
aspect_ratio: str,
|
||||
seed,
|
||||
style_image_weight: float,
|
||||
image_luma_ref: Optional[LumaReferenceChain] = None,
|
||||
style_image: Optional[torch.Tensor] = None,
|
||||
character_image: Optional[torch.Tensor] = None,
|
||||
image_luma_ref: LumaReferenceChain | None = None,
|
||||
style_image: torch.Tensor | None = None,
|
||||
character_image: torch.Tensor | None = None,
|
||||
) -> IO.NodeOutput:
|
||||
validate_string(prompt, strip_whitespace=True, min_length=3)
|
||||
# handle image_luma_ref
|
||||
@ -434,7 +436,7 @@ class LumaTextToVideoGenerationNode(IO.ComfyNode):
|
||||
duration: str,
|
||||
loop: bool,
|
||||
seed,
|
||||
luma_concepts: Optional[LumaConceptChain] = None,
|
||||
luma_concepts: LumaConceptChain | None = None,
|
||||
) -> IO.NodeOutput:
|
||||
validate_string(prompt, strip_whitespace=False, min_length=3)
|
||||
duration = duration if model != LumaVideoModel.ray_1_6 else None
|
||||
@ -533,7 +535,6 @@ class LumaImageToVideoGenerationNode(IO.ComfyNode):
|
||||
],
|
||||
is_api_node=True,
|
||||
price_badge=PRICE_BADGE_VIDEO,
|
||||
|
||||
)
|
||||
|
||||
@classmethod
|
||||
@ -644,6 +645,293 @@ PRICE_BADGE_VIDEO = IO.PriceBadge(
|
||||
)
|
||||
|
||||
|
||||
def _luma2_uni1_common_inputs(max_image_refs: int) -> list:
|
||||
return [
|
||||
IO.Combo.Input(
|
||||
"style",
|
||||
options=["auto", "manga"],
|
||||
default="auto",
|
||||
tooltip="Style preset. 'auto' picks based on the prompt; "
|
||||
"'manga' applies a manga/anime aesthetic and requires a portrait "
|
||||
"aspect ratio (2:3, 9:16, 1:2, 1:3).",
|
||||
),
|
||||
IO.Boolean.Input(
|
||||
"web_search",
|
||||
default=False,
|
||||
tooltip="Search the web for visual references before generating.",
|
||||
),
|
||||
IO.Autogrow.Input(
|
||||
"image_ref",
|
||||
template=IO.Autogrow.TemplateNames(
|
||||
IO.Image.Input("image"),
|
||||
names=[f"image_{i}" for i in range(1, max_image_refs + 1)],
|
||||
min=0,
|
||||
),
|
||||
optional=True,
|
||||
tooltip=f"Up to {max_image_refs} reference images for style/content guidance.",
|
||||
),
|
||||
]
|
||||
|
||||
|
||||
async def _luma2_upload_image_refs(
|
||||
cls: type[IO.ComfyNode],
|
||||
refs: dict | None,
|
||||
max_count: int,
|
||||
) -> list[Luma2ImageRef] | None:
|
||||
if not refs:
|
||||
return None
|
||||
out: list[Luma2ImageRef] = []
|
||||
for key in refs:
|
||||
url = await upload_image_to_comfyapi(cls, refs[key])
|
||||
out.append(Luma2ImageRef(url=url))
|
||||
if len(out) > max_count:
|
||||
raise ValueError(f"Maximum {max_count} reference images are allowed.")
|
||||
return out or None
|
||||
|
||||
|
||||
async def _luma2_submit_and_poll(
|
||||
cls: type[IO.ComfyNode],
|
||||
request: Luma2GenerationRequest,
|
||||
) -> Input.Image:
|
||||
initial = await sync_op(
|
||||
cls,
|
||||
ApiEndpoint(path="/proxy/luma_2/generations", method="POST"),
|
||||
response_model=Luma2Generation,
|
||||
data=request,
|
||||
)
|
||||
if not initial.id:
|
||||
raise RuntimeError("Luma 2 API did not return a generation id.")
|
||||
final = await poll_op(
|
||||
cls,
|
||||
ApiEndpoint(path=f"/proxy/luma_2/generations/{initial.id}", method="GET"),
|
||||
response_model=Luma2Generation,
|
||||
status_extractor=lambda r: r.state,
|
||||
progress_extractor=lambda r: None,
|
||||
)
|
||||
if not final.output:
|
||||
msg = final.failure_reason or "no output returned"
|
||||
raise RuntimeError(f"Luma 2 generation failed: {msg}")
|
||||
url = final.output[0].url
|
||||
if not url:
|
||||
raise RuntimeError("Luma 2 generation completed without an output URL.")
|
||||
return await download_url_to_image_tensor(url)
|
||||
|
||||
|
||||
class LumaImageNode(IO.ComfyNode):
|
||||
|
||||
@classmethod
|
||||
def define_schema(cls) -> IO.Schema:
|
||||
return IO.Schema(
|
||||
node_id="LumaImageNode2",
|
||||
display_name="Luma UNI-1 Image",
|
||||
category="api node/image/Luma",
|
||||
description="Generate images from text using the Luma UNI-1 model.",
|
||||
inputs=[
|
||||
IO.String.Input(
|
||||
"prompt",
|
||||
multiline=True,
|
||||
default="",
|
||||
tooltip="Text description of the desired image. 1–6000 characters.",
|
||||
),
|
||||
IO.DynamicCombo.Input(
|
||||
"model",
|
||||
options=[
|
||||
IO.DynamicCombo.Option(
|
||||
"uni-1",
|
||||
[
|
||||
IO.Combo.Input(
|
||||
"aspect_ratio",
|
||||
options=[
|
||||
"auto",
|
||||
"3:1",
|
||||
"2:1",
|
||||
"16:9",
|
||||
"3:2",
|
||||
"1:1",
|
||||
"2:3",
|
||||
"9:16",
|
||||
"1:2",
|
||||
"1:3",
|
||||
],
|
||||
default="auto",
|
||||
tooltip="Output image aspect ratio. 'auto' lets "
|
||||
"the model pick based on the prompt.",
|
||||
),
|
||||
*_luma2_uni1_common_inputs(max_image_refs=9),
|
||||
],
|
||||
),
|
||||
IO.DynamicCombo.Option(
|
||||
"uni-1-max",
|
||||
[
|
||||
IO.Combo.Input(
|
||||
"aspect_ratio",
|
||||
options=[
|
||||
"auto",
|
||||
"3:1",
|
||||
"2:1",
|
||||
"16:9",
|
||||
"3:2",
|
||||
"1:1",
|
||||
"2:3",
|
||||
"9:16",
|
||||
"1:2",
|
||||
"1:3",
|
||||
],
|
||||
default="auto",
|
||||
tooltip="Output image aspect ratio. 'auto' lets "
|
||||
"the model pick based on the prompt.",
|
||||
),
|
||||
*_luma2_uni1_common_inputs(max_image_refs=9),
|
||||
],
|
||||
),
|
||||
],
|
||||
tooltip="Model to use for generation.",
|
||||
),
|
||||
IO.Int.Input(
|
||||
"seed",
|
||||
default=0,
|
||||
min=0,
|
||||
max=2147483647,
|
||||
control_after_generate=True,
|
||||
tooltip="Seed controls whether the node should re-run; "
|
||||
"results are non-deterministic regardless of seed.",
|
||||
),
|
||||
],
|
||||
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,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["model"], input_groups=["model.image_ref"]),
|
||||
expr="""
|
||||
(
|
||||
$m := widgets.model;
|
||||
$refs := $lookup(inputGroups, "model.image_ref");
|
||||
$base := $m = "uni-1-max" ? 0.1 : 0.0404;
|
||||
{"type":"usd","usd": $round($base + 0.003 * $refs, 4)}
|
||||
)
|
||||
""",
|
||||
),
|
||||
)
|
||||
|
||||
@classmethod
|
||||
async def execute(
|
||||
cls,
|
||||
prompt: str,
|
||||
model: dict,
|
||||
seed: int,
|
||||
) -> IO.NodeOutput:
|
||||
validate_string(prompt, min_length=1, max_length=6000)
|
||||
aspect_ratio = model["aspect_ratio"]
|
||||
style = model["style"]
|
||||
allowed_manga_ratios = {"2:3", "9:16", "1:2", "1:3"}
|
||||
if style == "manga" and aspect_ratio != "auto" and aspect_ratio not in allowed_manga_ratios:
|
||||
raise ValueError(
|
||||
f"'manga' style requires a portrait aspect ratio "
|
||||
f"({', '.join(sorted(allowed_manga_ratios))}) or 'auto'; got '{aspect_ratio}'."
|
||||
)
|
||||
request = Luma2GenerationRequest(
|
||||
prompt=prompt,
|
||||
model=model["model"],
|
||||
type="image",
|
||||
aspect_ratio=aspect_ratio if aspect_ratio != "auto" else None,
|
||||
style=style if style != "auto" else None,
|
||||
output_format="png",
|
||||
web_search=model["web_search"],
|
||||
image_ref=await _luma2_upload_image_refs(cls, model.get("image_ref"), max_count=9),
|
||||
)
|
||||
return IO.NodeOutput(await _luma2_submit_and_poll(cls, request))
|
||||
|
||||
|
||||
class LumaImageEditNode(IO.ComfyNode):
|
||||
|
||||
@classmethod
|
||||
def define_schema(cls) -> IO.Schema:
|
||||
return IO.Schema(
|
||||
node_id="LumaImageEditNode2",
|
||||
display_name="Luma UNI-1 Image Edit",
|
||||
category="api node/image/Luma",
|
||||
description="Edit an existing image with a text prompt using the Luma UNI-1 model.",
|
||||
inputs=[
|
||||
IO.Image.Input(
|
||||
"source",
|
||||
tooltip="Source image to edit.",
|
||||
),
|
||||
IO.String.Input(
|
||||
"prompt",
|
||||
multiline=True,
|
||||
default="",
|
||||
tooltip="Description of the desired edit. 1–6000 characters.",
|
||||
),
|
||||
IO.DynamicCombo.Input(
|
||||
"model",
|
||||
options=[
|
||||
IO.DynamicCombo.Option(
|
||||
"uni-1",
|
||||
_luma2_uni1_common_inputs(max_image_refs=8),
|
||||
),
|
||||
IO.DynamicCombo.Option(
|
||||
"uni-1-max",
|
||||
_luma2_uni1_common_inputs(max_image_refs=8),
|
||||
),
|
||||
],
|
||||
tooltip="Model to use for editing.",
|
||||
),
|
||||
IO.Int.Input(
|
||||
"seed",
|
||||
default=0,
|
||||
min=0,
|
||||
max=2147483647,
|
||||
control_after_generate=True,
|
||||
tooltip="Seed controls whether the node should re-run; "
|
||||
"results are non-deterministic regardless of seed.",
|
||||
),
|
||||
],
|
||||
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,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(widgets=["model"], input_groups=["model.image_ref"]),
|
||||
expr="""
|
||||
(
|
||||
$m := widgets.model;
|
||||
$refs := $lookup(inputGroups, "model.image_ref");
|
||||
$base := $m = "uni-1-max" ? 0.103 : 0.0434;
|
||||
{"type":"usd","usd": $round($base + 0.003 * $refs, 4)}
|
||||
)
|
||||
""",
|
||||
),
|
||||
)
|
||||
|
||||
@classmethod
|
||||
async def execute(
|
||||
cls,
|
||||
source: Input.Image,
|
||||
prompt: str,
|
||||
model: dict,
|
||||
seed: int,
|
||||
) -> IO.NodeOutput:
|
||||
validate_string(prompt, min_length=1, max_length=6000)
|
||||
request = Luma2GenerationRequest(
|
||||
prompt=prompt,
|
||||
model=model["model"],
|
||||
type="image_edit",
|
||||
source=Luma2ImageRef(url=await upload_image_to_comfyapi(cls, source)),
|
||||
style=model["style"] if model["style"] != "auto" else None,
|
||||
output_format="png",
|
||||
web_search=model["web_search"],
|
||||
image_ref=await _luma2_upload_image_refs(cls, model.get("image_ref"), max_count=8),
|
||||
)
|
||||
return IO.NodeOutput(await _luma2_submit_and_poll(cls, request))
|
||||
|
||||
|
||||
class LumaExtension(ComfyExtension):
|
||||
@override
|
||||
async def get_node_list(self) -> list[type[IO.ComfyNode]]:
|
||||
@ -654,6 +942,8 @@ class LumaExtension(ComfyExtension):
|
||||
LumaImageToVideoGenerationNode,
|
||||
LumaReferenceNode,
|
||||
LumaConceptsNode,
|
||||
LumaImageNode,
|
||||
LumaImageEditNode,
|
||||
]
|
||||
|
||||
|
||||
|
||||
@ -230,7 +230,6 @@ class MagnificImageUpscalerCreativeNode(IO.ComfyNode):
|
||||
status_extractor=lambda x: x.status,
|
||||
price_extractor=lambda _: price_usd,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_image_tensor(final_response.generated[0]))
|
||||
|
||||
@ -391,7 +390,6 @@ class MagnificImageUpscalerPreciseV2Node(IO.ComfyNode):
|
||||
status_extractor=lambda x: x.status,
|
||||
price_extractor=lambda _: price_usd,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_image_tensor(final_response.generated[0]))
|
||||
|
||||
@ -541,7 +539,6 @@ class MagnificImageStyleTransferNode(IO.ComfyNode):
|
||||
response_model=TaskResponse,
|
||||
status_extractor=lambda x: x.status,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_image_tensor(final_response.generated[0]))
|
||||
|
||||
@ -782,7 +779,6 @@ class MagnificImageRelightNode(IO.ComfyNode):
|
||||
response_model=TaskResponse,
|
||||
status_extractor=lambda x: x.status,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_image_tensor(final_response.generated[0]))
|
||||
|
||||
@ -924,7 +920,6 @@ class MagnificImageSkinEnhancerNode(IO.ComfyNode):
|
||||
response_model=TaskResponse,
|
||||
status_extractor=lambda x: x.status,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_image_tensor(final_response.generated[0]))
|
||||
|
||||
|
||||
@ -1,534 +0,0 @@
|
||||
import logging
|
||||
|
||||
from typing_extensions import override
|
||||
|
||||
from comfy_api.latest import IO, ComfyExtension, Input
|
||||
from comfy_api_nodes.apis.moonvalley import (
|
||||
MoonvalleyPromptResponse,
|
||||
MoonvalleyTextToVideoInferenceParams,
|
||||
MoonvalleyTextToVideoRequest,
|
||||
MoonvalleyVideoToVideoInferenceParams,
|
||||
MoonvalleyVideoToVideoRequest,
|
||||
)
|
||||
from comfy_api_nodes.util import (
|
||||
ApiEndpoint,
|
||||
download_url_to_video_output,
|
||||
poll_op,
|
||||
sync_op,
|
||||
trim_video,
|
||||
upload_images_to_comfyapi,
|
||||
upload_video_to_comfyapi,
|
||||
validate_container_format_is_mp4,
|
||||
validate_image_dimensions,
|
||||
validate_string,
|
||||
)
|
||||
|
||||
API_UPLOADS_ENDPOINT = "/proxy/moonvalley/uploads"
|
||||
API_PROMPTS_ENDPOINT = "/proxy/moonvalley/prompts"
|
||||
API_VIDEO2VIDEO_ENDPOINT = "/proxy/moonvalley/prompts/video-to-video"
|
||||
API_TXT2VIDEO_ENDPOINT = "/proxy/moonvalley/prompts/text-to-video"
|
||||
API_IMG2VIDEO_ENDPOINT = "/proxy/moonvalley/prompts/image-to-video"
|
||||
|
||||
MIN_WIDTH = 300
|
||||
MIN_HEIGHT = 300
|
||||
|
||||
MAX_WIDTH = 10000
|
||||
MAX_HEIGHT = 10000
|
||||
|
||||
MIN_VID_WIDTH = 300
|
||||
MIN_VID_HEIGHT = 300
|
||||
|
||||
MAX_VID_WIDTH = 10000
|
||||
MAX_VID_HEIGHT = 10000
|
||||
|
||||
MAX_VIDEO_SIZE = 1024 * 1024 * 1024 # 1 GB max for in-memory video processing
|
||||
|
||||
MOONVALLEY_MAREY_MAX_PROMPT_LENGTH = 5000
|
||||
|
||||
|
||||
def is_valid_task_creation_response(response: MoonvalleyPromptResponse) -> bool:
|
||||
"""Verifies that the initial response contains a task ID."""
|
||||
return bool(response.id)
|
||||
|
||||
|
||||
def validate_task_creation_response(response) -> None:
|
||||
if not is_valid_task_creation_response(response):
|
||||
error_msg = f"Moonvalley Marey API: Initial request failed. Code: {response.code}, Message: {response.message}, Data: {response}"
|
||||
logging.error(error_msg)
|
||||
raise RuntimeError(error_msg)
|
||||
|
||||
|
||||
def validate_video_to_video_input(video: Input.Video) -> Input.Video:
|
||||
"""
|
||||
Validates and processes video input for Moonvalley Video-to-Video generation.
|
||||
|
||||
Args:
|
||||
video: Input video to validate
|
||||
|
||||
Returns:
|
||||
Validated and potentially trimmed video
|
||||
|
||||
Raises:
|
||||
ValueError: If video doesn't meet requirements
|
||||
MoonvalleyApiError: If video duration is too short
|
||||
"""
|
||||
width, height = _get_video_dimensions(video)
|
||||
_validate_video_dimensions(width, height)
|
||||
validate_container_format_is_mp4(video)
|
||||
|
||||
return _validate_and_trim_duration(video)
|
||||
|
||||
|
||||
def _get_video_dimensions(video: Input.Video) -> tuple[int, int]:
|
||||
"""Extracts video dimensions with error handling."""
|
||||
try:
|
||||
return video.get_dimensions()
|
||||
except Exception as e:
|
||||
logging.error("Error getting dimensions of video: %s", e)
|
||||
raise ValueError(f"Cannot get video dimensions: {e}") from e
|
||||
|
||||
|
||||
def _validate_video_dimensions(width: int, height: int) -> None:
|
||||
"""Validates video dimensions meet Moonvalley V2V requirements."""
|
||||
supported_resolutions = {
|
||||
(1920, 1080),
|
||||
(1080, 1920),
|
||||
(1152, 1152),
|
||||
(1536, 1152),
|
||||
(1152, 1536),
|
||||
}
|
||||
|
||||
if (width, height) not in supported_resolutions:
|
||||
supported_list = ", ".join([f"{w}x{h}" for w, h in sorted(supported_resolutions)])
|
||||
raise ValueError(f"Resolution {width}x{height} not supported. Supported: {supported_list}")
|
||||
|
||||
|
||||
def _validate_and_trim_duration(video: Input.Video) -> Input.Video:
|
||||
"""Validates video duration and trims to 5 seconds if needed."""
|
||||
duration = video.get_duration()
|
||||
_validate_minimum_duration(duration)
|
||||
return _trim_if_too_long(video, duration)
|
||||
|
||||
|
||||
def _validate_minimum_duration(duration: float) -> None:
|
||||
"""Ensures video is at least 5 seconds long."""
|
||||
if duration < 5:
|
||||
raise ValueError("Input video must be at least 5 seconds long.")
|
||||
|
||||
|
||||
def _trim_if_too_long(video: Input.Video, duration: float) -> Input.Video:
|
||||
"""Trims video to 5 seconds if longer."""
|
||||
if duration > 5:
|
||||
return trim_video(video, 5)
|
||||
return video
|
||||
|
||||
|
||||
def parse_width_height_from_res(resolution: str):
|
||||
# Accepts a string like "16:9 (1920 x 1080)" and returns width, height as a dict
|
||||
res_map = {
|
||||
"16:9 (1920 x 1080)": {"width": 1920, "height": 1080},
|
||||
"9:16 (1080 x 1920)": {"width": 1080, "height": 1920},
|
||||
"1:1 (1152 x 1152)": {"width": 1152, "height": 1152},
|
||||
"4:3 (1536 x 1152)": {"width": 1536, "height": 1152},
|
||||
"3:4 (1152 x 1536)": {"width": 1152, "height": 1536},
|
||||
# "21:9 (2560 x 1080)": {"width": 2560, "height": 1080},
|
||||
}
|
||||
return res_map.get(resolution, {"width": 1920, "height": 1080})
|
||||
|
||||
|
||||
def parse_control_parameter(value):
|
||||
control_map = {
|
||||
"Motion Transfer": "motion_control",
|
||||
"Canny": "canny_control",
|
||||
"Pose Transfer": "pose_control",
|
||||
"Depth": "depth_control",
|
||||
}
|
||||
return control_map.get(value, control_map["Motion Transfer"])
|
||||
|
||||
|
||||
async def get_response(cls: type[IO.ComfyNode], task_id: str) -> MoonvalleyPromptResponse:
|
||||
return await poll_op(
|
||||
cls,
|
||||
ApiEndpoint(path=f"{API_PROMPTS_ENDPOINT}/{task_id}"),
|
||||
response_model=MoonvalleyPromptResponse,
|
||||
status_extractor=lambda r: (r.status if r and r.status else None),
|
||||
poll_interval=16.0,
|
||||
max_poll_attempts=240,
|
||||
)
|
||||
|
||||
|
||||
class MoonvalleyImg2VideoNode(IO.ComfyNode):
|
||||
|
||||
@classmethod
|
||||
def define_schema(cls) -> IO.Schema:
|
||||
return IO.Schema(
|
||||
node_id="MoonvalleyImg2VideoNode",
|
||||
display_name="Moonvalley Marey Image to Video",
|
||||
category="api node/video/Moonvalley Marey",
|
||||
description="Moonvalley Marey Image to Video Node",
|
||||
inputs=[
|
||||
IO.Image.Input(
|
||||
"image",
|
||||
tooltip="The reference image used to generate the video",
|
||||
),
|
||||
IO.String.Input(
|
||||
"prompt",
|
||||
multiline=True,
|
||||
),
|
||||
IO.String.Input(
|
||||
"negative_prompt",
|
||||
multiline=True,
|
||||
default="<synthetic> <scene cut> gopro, bright, contrast, static, overexposed, vignette, "
|
||||
"artifacts, still, noise, texture, scanlines, videogame, 360 camera, VR, transition, "
|
||||
"flare, saturation, distorted, warped, wide angle, saturated, vibrant, glowing, "
|
||||
"cross dissolve, cheesy, ugly hands, mutated hands, mutant, disfigured, extra fingers, "
|
||||
"blown out, horrible, blurry, worst quality, bad, dissolve, melt, fade in, fade out, "
|
||||
"wobbly, weird, low quality, plastic, stock footage, video camera, boring",
|
||||
tooltip="Negative prompt text",
|
||||
),
|
||||
IO.Combo.Input(
|
||||
"resolution",
|
||||
options=[
|
||||
"16:9 (1920 x 1080)",
|
||||
"9:16 (1080 x 1920)",
|
||||
"1:1 (1152 x 1152)",
|
||||
"4:3 (1536 x 1152)",
|
||||
"3:4 (1152 x 1536)",
|
||||
# "21:9 (2560 x 1080)",
|
||||
],
|
||||
default="16:9 (1920 x 1080)",
|
||||
tooltip="Resolution of the output video",
|
||||
),
|
||||
IO.Float.Input(
|
||||
"prompt_adherence",
|
||||
default=4.5,
|
||||
min=1.0,
|
||||
max=20.0,
|
||||
step=1.0,
|
||||
tooltip="Guidance scale for generation control",
|
||||
),
|
||||
IO.Int.Input(
|
||||
"seed",
|
||||
default=9,
|
||||
min=0,
|
||||
max=4294967295,
|
||||
step=1,
|
||||
display_mode=IO.NumberDisplay.number,
|
||||
tooltip="Random seed value",
|
||||
control_after_generate=True,
|
||||
),
|
||||
IO.Int.Input(
|
||||
"steps",
|
||||
default=80,
|
||||
min=75, # steps should be greater or equal to cooldown_steps(75) + warmup_steps(0)
|
||||
max=100,
|
||||
step=1,
|
||||
tooltip="Number of denoising steps",
|
||||
),
|
||||
],
|
||||
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,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(),
|
||||
expr="""{"type":"usd","usd": 1.5}""",
|
||||
),
|
||||
)
|
||||
|
||||
@classmethod
|
||||
async def execute(
|
||||
cls,
|
||||
image: Input.Image,
|
||||
prompt: str,
|
||||
negative_prompt: str,
|
||||
resolution: str,
|
||||
prompt_adherence: float,
|
||||
seed: int,
|
||||
steps: int,
|
||||
) -> IO.NodeOutput:
|
||||
validate_image_dimensions(image, min_width=300, min_height=300, max_height=MAX_HEIGHT, max_width=MAX_WIDTH)
|
||||
validate_string(prompt, min_length=1, max_length=MOONVALLEY_MAREY_MAX_PROMPT_LENGTH)
|
||||
validate_string(negative_prompt, field_name="negative_prompt", max_length=MOONVALLEY_MAREY_MAX_PROMPT_LENGTH)
|
||||
width_height = parse_width_height_from_res(resolution)
|
||||
|
||||
inference_params = MoonvalleyTextToVideoInferenceParams(
|
||||
negative_prompt=negative_prompt,
|
||||
steps=steps,
|
||||
seed=seed,
|
||||
guidance_scale=prompt_adherence,
|
||||
width=width_height["width"],
|
||||
height=width_height["height"],
|
||||
use_negative_prompts=True,
|
||||
)
|
||||
|
||||
# Get MIME type from tensor - assuming PNG format for image tensors
|
||||
mime_type = "image/png"
|
||||
image_url = (await upload_images_to_comfyapi(cls, image, max_images=1, mime_type=mime_type))[0]
|
||||
task_creation_response = await sync_op(
|
||||
cls,
|
||||
endpoint=ApiEndpoint(path=API_IMG2VIDEO_ENDPOINT, method="POST"),
|
||||
response_model=MoonvalleyPromptResponse,
|
||||
data=MoonvalleyTextToVideoRequest(
|
||||
image_url=image_url, prompt_text=prompt, inference_params=inference_params
|
||||
),
|
||||
)
|
||||
validate_task_creation_response(task_creation_response)
|
||||
final_response = await get_response(cls, task_creation_response.id)
|
||||
video = await download_url_to_video_output(final_response.output_url)
|
||||
return IO.NodeOutput(video)
|
||||
|
||||
|
||||
class MoonvalleyVideo2VideoNode(IO.ComfyNode):
|
||||
|
||||
@classmethod
|
||||
def define_schema(cls) -> IO.Schema:
|
||||
return IO.Schema(
|
||||
node_id="MoonvalleyVideo2VideoNode",
|
||||
display_name="Moonvalley Marey Video to Video",
|
||||
category="api node/video/Moonvalley Marey",
|
||||
description="",
|
||||
inputs=[
|
||||
IO.String.Input(
|
||||
"prompt",
|
||||
multiline=True,
|
||||
tooltip="Describes the video to generate",
|
||||
),
|
||||
IO.String.Input(
|
||||
"negative_prompt",
|
||||
multiline=True,
|
||||
default="<synthetic> <scene cut> gopro, bright, contrast, static, overexposed, vignette, "
|
||||
"artifacts, still, noise, texture, scanlines, videogame, 360 camera, VR, transition, "
|
||||
"flare, saturation, distorted, warped, wide angle, saturated, vibrant, glowing, "
|
||||
"cross dissolve, cheesy, ugly hands, mutated hands, mutant, disfigured, extra fingers, "
|
||||
"blown out, horrible, blurry, worst quality, bad, dissolve, melt, fade in, fade out, "
|
||||
"wobbly, weird, low quality, plastic, stock footage, video camera, boring",
|
||||
tooltip="Negative prompt text",
|
||||
),
|
||||
IO.Int.Input(
|
||||
"seed",
|
||||
default=9,
|
||||
min=0,
|
||||
max=4294967295,
|
||||
step=1,
|
||||
display_mode=IO.NumberDisplay.number,
|
||||
tooltip="Random seed value",
|
||||
control_after_generate=False,
|
||||
),
|
||||
IO.Video.Input(
|
||||
"video",
|
||||
tooltip="The reference video used to generate the output video. Must be at least 5 seconds long. "
|
||||
"Videos longer than 5s will be automatically trimmed. Only MP4 format supported.",
|
||||
),
|
||||
IO.Combo.Input(
|
||||
"control_type",
|
||||
options=["Motion Transfer", "Pose Transfer"],
|
||||
default="Motion Transfer",
|
||||
optional=True,
|
||||
),
|
||||
IO.Int.Input(
|
||||
"motion_intensity",
|
||||
default=100,
|
||||
min=0,
|
||||
max=100,
|
||||
step=1,
|
||||
tooltip="Only used if control_type is 'Motion Transfer'",
|
||||
optional=True,
|
||||
),
|
||||
IO.Int.Input(
|
||||
"steps",
|
||||
default=60,
|
||||
min=60, # steps should be greater or equal to cooldown_steps(36) + warmup_steps(24)
|
||||
max=100,
|
||||
step=1,
|
||||
display_mode=IO.NumberDisplay.number,
|
||||
tooltip="Number of inference steps",
|
||||
),
|
||||
],
|
||||
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,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(),
|
||||
expr="""{"type":"usd","usd": 2.25}""",
|
||||
),
|
||||
)
|
||||
|
||||
@classmethod
|
||||
async def execute(
|
||||
cls,
|
||||
prompt: str,
|
||||
negative_prompt: str,
|
||||
seed: int,
|
||||
video: Input.Video | None = None,
|
||||
control_type: str = "Motion Transfer",
|
||||
motion_intensity: int | None = 100,
|
||||
steps=60,
|
||||
prompt_adherence=4.5,
|
||||
) -> IO.NodeOutput:
|
||||
validated_video = validate_video_to_video_input(video)
|
||||
video_url = await upload_video_to_comfyapi(cls, validated_video)
|
||||
validate_string(prompt, min_length=1, max_length=MOONVALLEY_MAREY_MAX_PROMPT_LENGTH)
|
||||
validate_string(negative_prompt, field_name="negative_prompt", max_length=MOONVALLEY_MAREY_MAX_PROMPT_LENGTH)
|
||||
|
||||
# Only include motion_intensity for Motion Transfer
|
||||
control_params = {}
|
||||
if control_type == "Motion Transfer" and motion_intensity is not None:
|
||||
control_params["motion_intensity"] = motion_intensity
|
||||
|
||||
inference_params = MoonvalleyVideoToVideoInferenceParams(
|
||||
negative_prompt=negative_prompt,
|
||||
seed=seed,
|
||||
control_params=control_params,
|
||||
steps=steps,
|
||||
guidance_scale=prompt_adherence,
|
||||
)
|
||||
|
||||
task_creation_response = await sync_op(
|
||||
cls,
|
||||
endpoint=ApiEndpoint(path=API_VIDEO2VIDEO_ENDPOINT, method="POST"),
|
||||
response_model=MoonvalleyPromptResponse,
|
||||
data=MoonvalleyVideoToVideoRequest(
|
||||
control_type=parse_control_parameter(control_type),
|
||||
video_url=video_url,
|
||||
prompt_text=prompt,
|
||||
inference_params=inference_params,
|
||||
),
|
||||
)
|
||||
validate_task_creation_response(task_creation_response)
|
||||
final_response = await get_response(cls, task_creation_response.id)
|
||||
return IO.NodeOutput(await download_url_to_video_output(final_response.output_url))
|
||||
|
||||
|
||||
class MoonvalleyTxt2VideoNode(IO.ComfyNode):
|
||||
|
||||
@classmethod
|
||||
def define_schema(cls) -> IO.Schema:
|
||||
return IO.Schema(
|
||||
node_id="MoonvalleyTxt2VideoNode",
|
||||
display_name="Moonvalley Marey Text to Video",
|
||||
category="api node/video/Moonvalley Marey",
|
||||
description="",
|
||||
inputs=[
|
||||
IO.String.Input(
|
||||
"prompt",
|
||||
multiline=True,
|
||||
),
|
||||
IO.String.Input(
|
||||
"negative_prompt",
|
||||
multiline=True,
|
||||
default="<synthetic> <scene cut> gopro, bright, contrast, static, overexposed, vignette, "
|
||||
"artifacts, still, noise, texture, scanlines, videogame, 360 camera, VR, transition, "
|
||||
"flare, saturation, distorted, warped, wide angle, saturated, vibrant, glowing, "
|
||||
"cross dissolve, cheesy, ugly hands, mutated hands, mutant, disfigured, extra fingers, "
|
||||
"blown out, horrible, blurry, worst quality, bad, dissolve, melt, fade in, fade out, "
|
||||
"wobbly, weird, low quality, plastic, stock footage, video camera, boring",
|
||||
tooltip="Negative prompt text",
|
||||
),
|
||||
IO.Combo.Input(
|
||||
"resolution",
|
||||
options=[
|
||||
"16:9 (1920 x 1080)",
|
||||
"9:16 (1080 x 1920)",
|
||||
"1:1 (1152 x 1152)",
|
||||
"4:3 (1536 x 1152)",
|
||||
"3:4 (1152 x 1536)",
|
||||
"21:9 (2560 x 1080)",
|
||||
],
|
||||
default="16:9 (1920 x 1080)",
|
||||
tooltip="Resolution of the output video",
|
||||
),
|
||||
IO.Float.Input(
|
||||
"prompt_adherence",
|
||||
default=4.0,
|
||||
min=1.0,
|
||||
max=20.0,
|
||||
step=1.0,
|
||||
tooltip="Guidance scale for generation control",
|
||||
),
|
||||
IO.Int.Input(
|
||||
"seed",
|
||||
default=9,
|
||||
min=0,
|
||||
max=4294967295,
|
||||
step=1,
|
||||
display_mode=IO.NumberDisplay.number,
|
||||
control_after_generate=True,
|
||||
tooltip="Random seed value",
|
||||
),
|
||||
IO.Int.Input(
|
||||
"steps",
|
||||
default=80,
|
||||
min=75, # steps should be greater or equal to cooldown_steps(75) + warmup_steps(0)
|
||||
max=100,
|
||||
step=1,
|
||||
tooltip="Inference steps",
|
||||
),
|
||||
],
|
||||
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,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(),
|
||||
expr="""{"type":"usd","usd": 1.5}""",
|
||||
),
|
||||
)
|
||||
|
||||
@classmethod
|
||||
async def execute(
|
||||
cls,
|
||||
prompt: str,
|
||||
negative_prompt: str,
|
||||
resolution: str,
|
||||
prompt_adherence: float,
|
||||
seed: int,
|
||||
steps: int,
|
||||
) -> IO.NodeOutput:
|
||||
validate_string(prompt, min_length=1, max_length=MOONVALLEY_MAREY_MAX_PROMPT_LENGTH)
|
||||
validate_string(negative_prompt, field_name="negative_prompt", max_length=MOONVALLEY_MAREY_MAX_PROMPT_LENGTH)
|
||||
width_height = parse_width_height_from_res(resolution)
|
||||
|
||||
inference_params = MoonvalleyTextToVideoInferenceParams(
|
||||
negative_prompt=negative_prompt,
|
||||
steps=steps,
|
||||
seed=seed,
|
||||
guidance_scale=prompt_adherence,
|
||||
num_frames=128,
|
||||
width=width_height["width"],
|
||||
height=width_height["height"],
|
||||
)
|
||||
|
||||
task_creation_response = await sync_op(
|
||||
cls,
|
||||
endpoint=ApiEndpoint(path=API_TXT2VIDEO_ENDPOINT, method="POST"),
|
||||
response_model=MoonvalleyPromptResponse,
|
||||
data=MoonvalleyTextToVideoRequest(prompt_text=prompt, inference_params=inference_params),
|
||||
)
|
||||
validate_task_creation_response(task_creation_response)
|
||||
final_response = await get_response(cls, task_creation_response.id)
|
||||
return IO.NodeOutput(await download_url_to_video_output(final_response.output_url))
|
||||
|
||||
|
||||
class MoonvalleyExtension(ComfyExtension):
|
||||
@override
|
||||
async def get_node_list(self) -> list[type[IO.ComfyNode]]:
|
||||
return [
|
||||
MoonvalleyImg2VideoNode,
|
||||
MoonvalleyTxt2VideoNode,
|
||||
MoonvalleyVideo2VideoNode,
|
||||
]
|
||||
|
||||
|
||||
async def comfy_entrypoint() -> MoonvalleyExtension:
|
||||
return MoonvalleyExtension()
|
||||
@ -39,16 +39,18 @@ STARTING_POINT_ID_PATTERN = r"<starting_point_id:(.*)>"
|
||||
|
||||
|
||||
class SupportedOpenAIModel(str, Enum):
|
||||
o4_mini = "o4-mini"
|
||||
o1 = "o1"
|
||||
o3 = "o3"
|
||||
o1_pro = "o1-pro"
|
||||
gpt_4_1 = "gpt-4.1"
|
||||
gpt_4_1_mini = "gpt-4.1-mini"
|
||||
gpt_4_1_nano = "gpt-4.1-nano"
|
||||
gpt_5_5_pro = "gpt-5.5-pro"
|
||||
gpt_5_5 = "gpt-5.5"
|
||||
gpt_5 = "gpt-5"
|
||||
gpt_5_mini = "gpt-5-mini"
|
||||
gpt_5_nano = "gpt-5-nano"
|
||||
gpt_4_1 = "gpt-4.1"
|
||||
gpt_4_1_mini = "gpt-4.1-mini"
|
||||
gpt_4_1_nano = "gpt-4.1-nano"
|
||||
o4_mini = "o4-mini"
|
||||
o3 = "o3"
|
||||
o1_pro = "o1-pro"
|
||||
o1 = "o1"
|
||||
|
||||
|
||||
async def validate_and_cast_response(response, timeout: int = None) -> torch.Tensor:
|
||||
@ -415,8 +417,9 @@ class OpenAIGPTImage1(IO.ComfyNode):
|
||||
"1152x2048",
|
||||
"3840x2160",
|
||||
"2160x3840",
|
||||
"Custom",
|
||||
],
|
||||
tooltip="Image size",
|
||||
tooltip="Image size. Select 'Custom' to use the custom width and height (GPT Image 2 only).",
|
||||
optional=True,
|
||||
),
|
||||
IO.Int.Input(
|
||||
@ -445,6 +448,24 @@ class OpenAIGPTImage1(IO.ComfyNode):
|
||||
default="gpt-image-2",
|
||||
optional=True,
|
||||
),
|
||||
IO.Int.Input(
|
||||
"custom_width",
|
||||
default=1024,
|
||||
min=1024,
|
||||
max=3840,
|
||||
step=16,
|
||||
tooltip="Used only when `size` is 'Custom'. Must be a multiple of 16 (GPT Image 2 only).",
|
||||
optional=True,
|
||||
),
|
||||
IO.Int.Input(
|
||||
"custom_height",
|
||||
default=1024,
|
||||
min=1024,
|
||||
max=3840,
|
||||
step=16,
|
||||
tooltip="Used only when `size` is 'Custom'. Must be a multiple of 16 (GPT Image 2 only).",
|
||||
optional=True,
|
||||
),
|
||||
],
|
||||
outputs=[
|
||||
IO.Image.Output(),
|
||||
@ -471,9 +492,9 @@ class OpenAIGPTImage1(IO.ComfyNode):
|
||||
"high": [0.133, 0.22]
|
||||
},
|
||||
"gpt-image-2": {
|
||||
"low": [0.0048, 0.012],
|
||||
"medium": [0.041, 0.112],
|
||||
"high": [0.165, 0.43]
|
||||
"low": [0.0048, 0.019],
|
||||
"medium": [0.041, 0.168],
|
||||
"high": [0.165, 0.67]
|
||||
}
|
||||
};
|
||||
$range := $lookup($lookup($ranges, widgets.model), widgets.quality);
|
||||
@ -503,6 +524,8 @@ class OpenAIGPTImage1(IO.ComfyNode):
|
||||
mask: Input.Image | None = None,
|
||||
n: int = 1,
|
||||
size: str = "1024x1024",
|
||||
custom_width: int = 1024,
|
||||
custom_height: int = 1024,
|
||||
model: str = "gpt-image-1",
|
||||
) -> IO.NodeOutput:
|
||||
validate_string(prompt, strip_whitespace=False)
|
||||
@ -510,7 +533,25 @@ class OpenAIGPTImage1(IO.ComfyNode):
|
||||
if mask is not None and image is None:
|
||||
raise ValueError("Cannot use a mask without an input image")
|
||||
|
||||
if model in ("gpt-image-1", "gpt-image-1.5"):
|
||||
if size == "Custom":
|
||||
if model != "gpt-image-2":
|
||||
raise ValueError("Custom resolution is only supported by GPT Image 2 model")
|
||||
if custom_width % 16 != 0 or custom_height % 16 != 0:
|
||||
raise ValueError(f"Custom width and height must be multiples of 16, got {custom_width}x{custom_height}")
|
||||
if max(custom_width, custom_height) > 3840:
|
||||
raise ValueError(f"Custom resolution max edge must be <= 3840, got {custom_width}x{custom_height}")
|
||||
ratio = max(custom_width, custom_height) / min(custom_width, custom_height)
|
||||
if ratio > 3:
|
||||
raise ValueError(
|
||||
f"Custom resolution aspect ratio must not exceed 3:1, got {custom_width}x{custom_height}"
|
||||
)
|
||||
total_pixels = custom_width * custom_height
|
||||
if not 655_360 <= total_pixels <= 8_294_400:
|
||||
raise ValueError(
|
||||
f"Custom resolution total pixels must be between 655,360 and 8,294,400, got {total_pixels}"
|
||||
)
|
||||
size = f"{custom_width}x{custom_height}"
|
||||
elif model in ("gpt-image-1", "gpt-image-1.5"):
|
||||
if size not in ("auto", "1024x1024", "1024x1536", "1536x1024"):
|
||||
raise ValueError(f"Resolution {size} is only supported by GPT Image 2 model")
|
||||
|
||||
@ -700,6 +741,16 @@ class OpenAIChatNode(IO.ComfyNode):
|
||||
"usd": [0.002, 0.008],
|
||||
"format": { "approximate": true, "separator": "-", "suffix": " per 1K tokens" }
|
||||
}
|
||||
: $contains($m, "gpt-5.5-pro") ? {
|
||||
"type": "list_usd",
|
||||
"usd": [0.03, 0.18],
|
||||
"format": { "approximate": true, "separator": "-", "suffix": " per 1K tokens" }
|
||||
}
|
||||
: $contains($m, "gpt-5.5") ? {
|
||||
"type": "list_usd",
|
||||
"usd": [0.005, 0.03],
|
||||
"format": { "approximate": true, "separator": "-", "suffix": " per 1K tokens" }
|
||||
}
|
||||
: $contains($m, "gpt-5-nano") ? {
|
||||
"type": "list_usd",
|
||||
"usd": [0.00005, 0.0004],
|
||||
|
||||
@ -36,11 +36,15 @@ from comfy_api_nodes.util import (
|
||||
)
|
||||
|
||||
UPSCALER_MODELS_MAP = {
|
||||
"Astra 2": "ast-2",
|
||||
"Starlight (Astra) Fast": "slf-1",
|
||||
"Starlight (Astra) Creative": "slc-1",
|
||||
"Starlight Precise 2.5": "slp-2.5",
|
||||
}
|
||||
|
||||
AST2_MAX_FRAMES = 9000
|
||||
AST2_MAX_FRAMES_WITH_PROMPT = 450
|
||||
|
||||
|
||||
class TopazImageEnhance(IO.ComfyNode):
|
||||
@classmethod
|
||||
@ -230,13 +234,20 @@ class TopazVideoEnhance(IO.ComfyNode):
|
||||
def define_schema(cls):
|
||||
return IO.Schema(
|
||||
node_id="TopazVideoEnhance",
|
||||
display_name="Topaz Video Enhance",
|
||||
display_name="Topaz Video Enhance (Legacy)",
|
||||
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_model",
|
||||
options=[
|
||||
"Starlight (Astra) Fast",
|
||||
"Starlight (Astra) Creative",
|
||||
"Starlight Precise 2.5",
|
||||
],
|
||||
),
|
||||
IO.Combo.Input("upscaler_resolution", options=["FullHD (1080p)", "4K (2160p)"]),
|
||||
IO.Combo.Input(
|
||||
"upscaler_creativity",
|
||||
@ -304,6 +315,7 @@ class TopazVideoEnhance(IO.ComfyNode):
|
||||
IO.Hidden.unique_id,
|
||||
],
|
||||
is_api_node=True,
|
||||
is_deprecated=True,
|
||||
)
|
||||
|
||||
@classmethod
|
||||
@ -453,7 +465,350 @@ class TopazVideoEnhance(IO.ComfyNode):
|
||||
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 TopazVideoEnhanceV2(IO.ComfyNode):
|
||||
@classmethod
|
||||
def define_schema(cls):
|
||||
return IO.Schema(
|
||||
node_id="TopazVideoEnhanceV2",
|
||||
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.DynamicCombo.Input(
|
||||
"upscaler_model",
|
||||
options=[
|
||||
IO.DynamicCombo.Option(
|
||||
"Astra 2",
|
||||
[
|
||||
IO.Combo.Input("upscaler_resolution", options=["FullHD (1080p)", "4K (2160p)"]),
|
||||
IO.Float.Input(
|
||||
"creativity",
|
||||
default=0.5,
|
||||
min=0.0,
|
||||
max=1.0,
|
||||
step=0.1,
|
||||
display_mode=IO.NumberDisplay.slider,
|
||||
tooltip="Creative strength of the upscale.",
|
||||
),
|
||||
IO.String.Input(
|
||||
"prompt",
|
||||
multiline=True,
|
||||
default="",
|
||||
tooltip="Optional descriptive (not instructive) scene prompt."
|
||||
f"Capping input at {AST2_MAX_FRAMES_WITH_PROMPT} frames (~15s @ 30fps) when set.",
|
||||
),
|
||||
IO.Float.Input(
|
||||
"sharp",
|
||||
default=0.5,
|
||||
min=0.0,
|
||||
max=1.0,
|
||||
step=0.01,
|
||||
display_mode=IO.NumberDisplay.slider,
|
||||
tooltip="Pre-enhance sharpness: "
|
||||
"0.0=Gaussian blur, 0.5=passthrough (default), 1.0=USM sharpening.",
|
||||
advanced=True,
|
||||
),
|
||||
IO.Float.Input(
|
||||
"realism",
|
||||
default=0.0,
|
||||
min=0.0,
|
||||
max=1.0,
|
||||
step=0.01,
|
||||
display_mode=IO.NumberDisplay.slider,
|
||||
tooltip="Pulls output toward photographic realism."
|
||||
"Leave at 0 for the model default.",
|
||||
advanced=True,
|
||||
),
|
||||
],
|
||||
),
|
||||
IO.DynamicCombo.Option(
|
||||
"Starlight (Astra) Fast",
|
||||
[IO.Combo.Input("upscaler_resolution", options=["FullHD (1080p)", "4K (2160p)"]),],
|
||||
),
|
||||
IO.DynamicCombo.Option(
|
||||
"Starlight (Astra) Creative",
|
||||
[
|
||||
IO.Combo.Input("upscaler_resolution", options=["FullHD (1080p)", "4K (2160p)"]),
|
||||
IO.Combo.Input(
|
||||
"creativity",
|
||||
options=["low", "middle", "high"],
|
||||
default="low",
|
||||
tooltip="Creative strength of the upscale.",
|
||||
),
|
||||
],
|
||||
),
|
||||
IO.DynamicCombo.Option(
|
||||
"Starlight Precise 2.5",
|
||||
[IO.Combo.Input("upscaler_resolution", options=["FullHD (1080p)", "4K (2160p)"])],
|
||||
),
|
||||
IO.DynamicCombo.Option("Disabled", []),
|
||||
],
|
||||
),
|
||||
IO.DynamicCombo.Input(
|
||||
"interpolation_model",
|
||||
options=[
|
||||
IO.DynamicCombo.Option("Disabled", []),
|
||||
IO.DynamicCombo.Option(
|
||||
"apo-8",
|
||||
[
|
||||
IO.Int.Input(
|
||||
"interpolation_frame_rate",
|
||||
default=60,
|
||||
min=15,
|
||||
max=240,
|
||||
display_mode=IO.NumberDisplay.number,
|
||||
tooltip="Output frame rate.",
|
||||
),
|
||||
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.",
|
||||
advanced=True,
|
||||
),
|
||||
IO.Boolean.Input(
|
||||
"interpolation_duplicate",
|
||||
default=False,
|
||||
tooltip="Analyze the input for duplicate frames and remove them.",
|
||||
advanced=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.",
|
||||
advanced=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,
|
||||
price_badge=IO.PriceBadge(
|
||||
depends_on=IO.PriceBadgeDepends(widgets=[
|
||||
"upscaler_model",
|
||||
"upscaler_model.upscaler_resolution",
|
||||
"interpolation_model",
|
||||
]),
|
||||
expr="""
|
||||
(
|
||||
$model := $lookup(widgets, "upscaler_model");
|
||||
$res := $lookup(widgets, "upscaler_model.upscaler_resolution");
|
||||
$interp := $lookup(widgets, "interpolation_model");
|
||||
$is4k := $contains($res, "4k");
|
||||
$hasInterp := $interp != "disabled";
|
||||
$rates := {
|
||||
"starlight (astra) fast": {"hd": 0.43, "uhd": 0.85},
|
||||
"starlight precise 2.5": {"hd": 0.70, "uhd": 1.54},
|
||||
"astra 2": {"hd": 1.72, "uhd": 2.85},
|
||||
"starlight (astra) creative": {"hd": 2.25, "uhd": 3.99}
|
||||
};
|
||||
$surcharge := $is4k ? 0.28 : 0.14;
|
||||
$entry := $lookup($rates, $model);
|
||||
$base := $is4k ? $entry.uhd : $entry.hd;
|
||||
$hi := $base + ($hasInterp ? $surcharge : 0);
|
||||
$model = "disabled"
|
||||
? {"type":"text","text":"Interpolation only"}
|
||||
: ($hasInterp
|
||||
? {"type":"text","text":"~" & $string($base) & "–" & $string($hi) & " credits/src frame"}
|
||||
: {"type":"text","text":"~" & $string($base) & " credits/src frame"})
|
||||
)
|
||||
""",
|
||||
),
|
||||
)
|
||||
|
||||
@classmethod
|
||||
async def execute(
|
||||
cls,
|
||||
video: Input.Video,
|
||||
upscaler_model: dict,
|
||||
interpolation_model: dict,
|
||||
dynamic_compression_level: str = "Low",
|
||||
) -> IO.NodeOutput:
|
||||
upscaler_choice = upscaler_model["upscaler_model"]
|
||||
interpolation_choice = interpolation_model["interpolation_model"]
|
||||
if upscaler_choice == "Disabled" and interpolation_choice == "Disabled":
|
||||
raise ValueError("There is nothing to do: both upscaling and interpolation are disabled.")
|
||||
validate_container_format_is_mp4(video)
|
||||
src_width, src_height = video.get_dimensions()
|
||||
src_frame_rate = int(video.get_frame_rate())
|
||||
duration_sec = video.get_duration()
|
||||
src_video_stream = video.get_stream_source()
|
||||
target_width = src_width
|
||||
target_height = src_height
|
||||
target_frame_rate = src_frame_rate
|
||||
filters = []
|
||||
if upscaler_choice != "Disabled":
|
||||
if "1080p" in upscaler_model["upscaler_resolution"]:
|
||||
target_pixel_p = 1080
|
||||
max_long_side = 1920
|
||||
else:
|
||||
target_pixel_p = 2160
|
||||
max_long_side = 3840
|
||||
ar = src_width / src_height
|
||||
if src_width >= src_height:
|
||||
# Landscape or Square; Attempt to set height to target (e.g., 2160), calculate width
|
||||
target_height = target_pixel_p
|
||||
target_width = int(target_height * ar)
|
||||
# Check if width exceeds standard bounds (for ultra-wide e.g., 21:9 ARs)
|
||||
if target_width > max_long_side:
|
||||
target_width = max_long_side
|
||||
target_height = int(target_width / ar)
|
||||
else:
|
||||
# Portrait; Attempt to set width to target (e.g., 2160), calculate height
|
||||
target_width = target_pixel_p
|
||||
target_height = int(target_width / ar)
|
||||
# Check if height exceeds standard bounds
|
||||
if target_height > max_long_side:
|
||||
target_height = max_long_side
|
||||
target_width = int(target_height * ar)
|
||||
if target_width % 2 != 0:
|
||||
target_width += 1
|
||||
if target_height % 2 != 0:
|
||||
target_height += 1
|
||||
model_id = UPSCALER_MODELS_MAP[upscaler_choice]
|
||||
if model_id == "slc-1":
|
||||
filters.append(
|
||||
VideoEnhancementFilter(
|
||||
model=model_id,
|
||||
creativity=upscaler_model["creativity"],
|
||||
isOptimizedMode=True,
|
||||
)
|
||||
)
|
||||
elif model_id == "ast-2":
|
||||
n_frames = video.get_frame_count()
|
||||
ast2_prompt = (upscaler_model["prompt"] or "").strip()
|
||||
if ast2_prompt and n_frames > AST2_MAX_FRAMES_WITH_PROMPT:
|
||||
raise ValueError(
|
||||
f"Astra 2 with a prompt is limited to {AST2_MAX_FRAMES_WITH_PROMPT} input frames "
|
||||
f"(~15s @ 30fps); video has {n_frames}. Clear the prompt or shorten the clip."
|
||||
)
|
||||
if n_frames > AST2_MAX_FRAMES:
|
||||
raise ValueError(f"Astra 2 is limited to {AST2_MAX_FRAMES} input frames; video has {n_frames}.")
|
||||
realism = upscaler_model["realism"]
|
||||
filters.append(
|
||||
VideoEnhancementFilter(
|
||||
model=model_id,
|
||||
creativity=upscaler_model["creativity"],
|
||||
prompt=(ast2_prompt or None),
|
||||
sharp=upscaler_model["sharp"],
|
||||
realism=(realism if realism > 0 else None),
|
||||
)
|
||||
)
|
||||
else:
|
||||
filters.append(VideoEnhancementFilter(model=model_id))
|
||||
if interpolation_choice != "Disabled":
|
||||
target_frame_rate = interpolation_model["interpolation_frame_rate"]
|
||||
filters.append(
|
||||
VideoFrameInterpolationFilter(
|
||||
model=interpolation_choice,
|
||||
slowmo=interpolation_model["interpolation_slowmo"],
|
||||
fps=interpolation_model["interpolation_frame_rate"],
|
||||
duplicate=interpolation_model["interpolation_duplicate"],
|
||||
duplicate_threshold=interpolation_model["interpolation_duplicate_threshold"],
|
||||
),
|
||||
)
|
||||
initial_res = await sync_op(
|
||||
cls,
|
||||
ApiEndpoint(path="/proxy/topaz/video/", method="POST"),
|
||||
response_model=CreateVideoResponse,
|
||||
data=CreateVideoRequest(
|
||||
source=CreateVideoRequestSource(
|
||||
container="mp4",
|
||||
size=get_fs_object_size(src_video_stream),
|
||||
duration=int(duration_sec),
|
||||
frameCount=video.get_frame_count(),
|
||||
frameRate=src_frame_rate,
|
||||
resolution=Resolution(width=src_width, height=src_height),
|
||||
),
|
||||
filters=filters,
|
||||
output=OutputInformationVideo(
|
||||
resolution=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=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=VideoCompleteUploadResponse,
|
||||
data=VideoCompleteUploadRequest(
|
||||
uploadResults=[
|
||||
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=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,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(final_response.download.url))
|
||||
|
||||
@ -464,6 +819,7 @@ class TopazExtension(ComfyExtension):
|
||||
return [
|
||||
TopazImageEnhance,
|
||||
TopazVideoEnhance,
|
||||
TopazVideoEnhanceV2,
|
||||
]
|
||||
|
||||
|
||||
|
||||
@ -38,7 +38,7 @@ async def execute_task(
|
||||
cls: type[IO.ComfyNode],
|
||||
vidu_endpoint: str,
|
||||
payload: TaskCreationRequest | TaskExtendCreationRequest | TaskMultiFrameCreationRequest,
|
||||
max_poll_attempts: int = 320,
|
||||
max_poll_attempts: int = 480,
|
||||
) -> list[TaskResult]:
|
||||
task_creation_response = await sync_op(
|
||||
cls,
|
||||
@ -1097,7 +1097,6 @@ class ViduExtendVideoNode(IO.ComfyNode):
|
||||
video_url=await upload_video_to_comfyapi(cls, video, wait_label="Uploading video"),
|
||||
images=[image_url] if image_url else None,
|
||||
),
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(results[0].url))
|
||||
|
||||
|
||||
@ -818,7 +818,6 @@ class WanReferenceVideoApi(IO.ComfyNode):
|
||||
response_model=VideoTaskStatusResponse,
|
||||
status_extractor=lambda x: x.output.task_status,
|
||||
poll_interval=6,
|
||||
max_poll_attempts=280,
|
||||
)
|
||||
return IO.NodeOutput(await download_url_to_video_output(response.output.video_url))
|
||||
|
||||
|
||||
@ -84,7 +84,6 @@ class WavespeedFlashVSRNode(IO.ComfyNode):
|
||||
response_model=TaskResultResponse,
|
||||
status_extractor=lambda x: "failed" if x.data is None else x.data.status,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
if final_response.code != 200:
|
||||
raise ValueError(
|
||||
@ -156,7 +155,6 @@ class WavespeedImageUpscaleNode(IO.ComfyNode):
|
||||
response_model=TaskResultResponse,
|
||||
status_extractor=lambda x: "failed" if x.data is None else x.data.status,
|
||||
poll_interval=10.0,
|
||||
max_poll_attempts=480,
|
||||
)
|
||||
if final_response.code != 200:
|
||||
raise ValueError(
|
||||
|
||||
@ -148,7 +148,7 @@ async def poll_op(
|
||||
queued_statuses: list[str | int] | None = None,
|
||||
data: BaseModel | None = None,
|
||||
poll_interval: float = 5.0,
|
||||
max_poll_attempts: int = 160,
|
||||
max_poll_attempts: int = 480,
|
||||
timeout_per_poll: float = 120.0,
|
||||
max_retries_per_poll: int = 10,
|
||||
retry_delay_per_poll: float = 1.0,
|
||||
@ -254,7 +254,7 @@ async def poll_op_raw(
|
||||
queued_statuses: list[str | int] | None = None,
|
||||
data: dict[str, Any] | BaseModel | None = None,
|
||||
poll_interval: float = 5.0,
|
||||
max_poll_attempts: int = 160,
|
||||
max_poll_attempts: int = 480,
|
||||
timeout_per_poll: float = 120.0,
|
||||
max_retries_per_poll: int = 10,
|
||||
retry_delay_per_poll: float = 1.0,
|
||||
|
||||
@ -459,27 +459,23 @@ class SDPoseKeypointExtractor(io.ComfyNode):
|
||||
total_images = image.shape[0]
|
||||
captured_feat = None
|
||||
|
||||
model_h = int(head.heatmap_size[0]) * 4 # e.g. 192 * 4 = 768
|
||||
model_w = int(head.heatmap_size[1]) * 4 # e.g. 256 * 4 = 1024
|
||||
model_w = int(head.heatmap_size[0]) * 4 # 192 * 4 = 768
|
||||
model_h = int(head.heatmap_size[1]) * 4 # 256 * 4 = 1024
|
||||
|
||||
def _resize_to_model(imgs):
|
||||
"""Aspect-preserving resize + zero-pad BHWC images to (model_h, model_w). Returns (resized_bhwc, scale, pad_top, pad_left)."""
|
||||
"""Stretch BHWC images to (model_h, model_w), model expects no aspect preservation."""
|
||||
h, w = imgs.shape[-3], imgs.shape[-2]
|
||||
scale = min(model_h / h, model_w / w)
|
||||
sh, sw = int(round(h * scale)), int(round(w * scale))
|
||||
pt, pl = (model_h - sh) // 2, (model_w - sw) // 2
|
||||
method = "area" if (model_h <= h and model_w <= w) else "bilinear"
|
||||
chw = imgs.permute(0, 3, 1, 2).float()
|
||||
scaled = comfy.utils.common_upscale(chw, sw, sh, upscale_method="bilinear", crop="disabled")
|
||||
padded = torch.zeros(scaled.shape[0], scaled.shape[1], model_h, model_w, dtype=scaled.dtype, device=scaled.device)
|
||||
padded[:, :, pt:pt + sh, pl:pl + sw] = scaled
|
||||
return padded.permute(0, 2, 3, 1), scale, pt, pl
|
||||
scaled = comfy.utils.common_upscale(chw, model_w, model_h, upscale_method=method, crop="disabled")
|
||||
return scaled.permute(0, 2, 3, 1), model_w / w, model_h / h
|
||||
|
||||
def _remap_keypoints(kp, scale, pad_top, pad_left, offset_x=0, offset_y=0):
|
||||
def _remap_keypoints(kp, scale_x, scale_y, offset_x=0, offset_y=0):
|
||||
"""Remap keypoints from model space back to original image space."""
|
||||
kp = kp.copy() if isinstance(kp, np.ndarray) else np.array(kp, dtype=np.float32)
|
||||
invalid = kp[..., 0] < 0
|
||||
kp[..., 0] = (kp[..., 0] - pad_left) / scale + offset_x
|
||||
kp[..., 1] = (kp[..., 1] - pad_top) / scale + offset_y
|
||||
kp[..., 0] = kp[..., 0] / scale_x + offset_x
|
||||
kp[..., 1] = kp[..., 1] / scale_y + offset_y
|
||||
kp[invalid] = -1
|
||||
return kp
|
||||
|
||||
@ -529,18 +525,18 @@ class SDPoseKeypointExtractor(io.ComfyNode):
|
||||
continue
|
||||
|
||||
crop = img[:, y1:y2, x1:x2, :] # (1, crop_h, crop_w, C)
|
||||
crop_resized, scale, pad_top, pad_left = _resize_to_model(crop)
|
||||
crop_resized, sx, sy = _resize_to_model(crop)
|
||||
|
||||
latent_crop = vae.encode(crop_resized)
|
||||
kp_batch, sc_batch = _run_on_latent(latent_crop)
|
||||
kp = _remap_keypoints(kp_batch[0], scale, pad_top, pad_left, x1, y1)
|
||||
kp = _remap_keypoints(kp_batch[0], sx, sy, x1, y1)
|
||||
img_keypoints.append(kp)
|
||||
img_scores.append(sc_batch[0])
|
||||
else:
|
||||
img_resized, scale, pad_top, pad_left = _resize_to_model(img)
|
||||
img_resized, sx, sy = _resize_to_model(img)
|
||||
latent_img = vae.encode(img_resized)
|
||||
kp_batch, sc_batch = _run_on_latent(latent_img)
|
||||
img_keypoints.append(_remap_keypoints(kp_batch[0], scale, pad_top, pad_left))
|
||||
img_keypoints.append(_remap_keypoints(kp_batch[0], sx, sy))
|
||||
img_scores.append(sc_batch[0])
|
||||
|
||||
all_keypoints.append(img_keypoints)
|
||||
@ -549,12 +545,12 @@ class SDPoseKeypointExtractor(io.ComfyNode):
|
||||
|
||||
else: # full-image mode, batched
|
||||
for batch_start in tqdm(range(0, total_images, batch_size), desc="Extracting keypoints"):
|
||||
batch_resized, scale, pad_top, pad_left = _resize_to_model(image[batch_start:batch_start + batch_size])
|
||||
batch_resized, sx, sy = _resize_to_model(image[batch_start:batch_start + batch_size])
|
||||
latent_batch = vae.encode(batch_resized)
|
||||
kp_batch, sc_batch = _run_on_latent(latent_batch)
|
||||
|
||||
for kp, sc in zip(kp_batch, sc_batch):
|
||||
all_keypoints.append([_remap_keypoints(kp, scale, pad_top, pad_left)])
|
||||
all_keypoints.append([_remap_keypoints(kp, sx, sy)])
|
||||
all_scores.append([sc])
|
||||
|
||||
pbar.update(len(kp_batch))
|
||||
@ -727,13 +723,13 @@ class CropByBBoxes(io.ComfyNode):
|
||||
scale = min(output_width / crop_w, output_height / crop_h)
|
||||
scaled_w = int(round(crop_w * scale))
|
||||
scaled_h = int(round(crop_h * scale))
|
||||
scaled = comfy.utils.common_upscale(crop_chw, scaled_w, scaled_h, upscale_method="bilinear", crop="disabled")
|
||||
scaled = comfy.utils.common_upscale(crop_chw, scaled_w, scaled_h, upscale_method="area", crop="disabled")
|
||||
pad_left = (output_width - scaled_w) // 2
|
||||
pad_top = (output_height - scaled_h) // 2
|
||||
resized = torch.zeros(1, num_ch, output_height, output_width, dtype=image.dtype, device=image.device)
|
||||
resized[:, :, pad_top:pad_top + scaled_h, pad_left:pad_left + scaled_w] = scaled
|
||||
else: # "stretch"
|
||||
resized = comfy.utils.common_upscale(crop_chw, output_width, output_height, upscale_method="bilinear", crop="disabled")
|
||||
resized = comfy.utils.common_upscale(crop_chw, output_width, output_height, upscale_method="area", crop="disabled")
|
||||
crops.append(resized)
|
||||
|
||||
if not crops:
|
||||
|
||||
@ -1,3 +1,3 @@
|
||||
# This file is automatically generated by the build process when version is
|
||||
# updated in pyproject.toml.
|
||||
__version__ = "0.20.1"
|
||||
__version__ = "0.20.3"
|
||||
|
||||
@ -86,6 +86,6 @@ def image_alpha_fix(destination, source):
|
||||
if destination.shape[-1] < source.shape[-1]:
|
||||
source = source[...,:destination.shape[-1]]
|
||||
elif destination.shape[-1] > source.shape[-1]:
|
||||
destination = torch.nn.functional.pad(destination, (0, 1))
|
||||
destination[..., -1] = 1.0
|
||||
source = torch.nn.functional.pad(source, (0, 1))
|
||||
source[..., -1] = 1.0
|
||||
return destination, source
|
||||
|
||||
@ -1,6 +1,6 @@
|
||||
[project]
|
||||
name = "ComfyUI"
|
||||
version = "0.20.1"
|
||||
version = "0.20.3"
|
||||
readme = "README.md"
|
||||
license = { file = "LICENSE" }
|
||||
requires-python = ">=3.10"
|
||||
|
||||
@ -1,5 +1,5 @@
|
||||
comfyui-frontend-package==1.42.15
|
||||
comfyui-workflow-templates==0.9.63
|
||||
comfyui-workflow-templates==0.9.72
|
||||
comfyui-embedded-docs==0.4.4
|
||||
torch
|
||||
torchsde
|
||||
|
||||
Reference in New Issue
Block a user