n8n (version 1.0.0 or later)
ComfyUI instance running and accessible
Node.js 16 or newer
For video generation: ComfyUI with AnimateDiff, SVD, or other video generation extensions
For audio generation: ComfyUI with ACE Step Audio models and extensions

Installation

Standard Installation

# Install from npm package (recommended)
npm install n8n-nodes-comfyui-image-to-image@1.0.0

# Or install from git repository
npm install github:StudioExitt/n8n-nodes-comfyui-image-to-image

Docker Installation

For Docker installations, you can use one of the following methods:

Method 1: Volume Mount

# Create a directory for custom nodes
mkdir -p /path/to/custom/nodes

# Clone the repository
git clone https://github.com/StudioExitt/n8n-nodes-comfyui-image-to-image.git /tmp/custom-nodes
cd /tmp/custom-nodes

# Install dependencies and build
npm install
npm run build

# Copy the built files to your custom nodes directory
cp -r dist/* /path/to/custom/nodes/n8n-nodes-comfyui-image-to-image/

# Run n8n with the custom nodes directory mounted
docker run -d \
  --name n8n \
  -p 5678:5678 \
  -v /path/to/custom/nodes:/home/node/.n8n/custom/nodes \
  n8nio/n8n

Method 2: Custom Docker Image

# Dockerfile
FROM n8nio/n8n

# Install the custom nodes package from GitHub
RUN cd /tmp && \
    npm install github:StudioExitt/n8n-nodes-comfyui-image-to-image && \
    cp -r /tmp/node_modules/n8n-nodes-comfyui-image-to-image /home/node/.n8n/custom/nodes/ && \
    chown -R node:node /home/node/.n8n

Build and run:

docker build -t n8n-with-comfyui .
docker run -d --name n8n -p 5678:5678 n8n-with-comfyui

Method 3: Docker Compose

version: '3'

services:
  n8n:
    image: n8nio/n8n
    container_name: n8n
    ports:
      - "5678:5678"
    volumes:
      - n8n_data:/home/node/.n8n
      - ./custom-nodes:/home/node/.n8n/custom/nodes
    environment:
      - N8N_PORT=5678
      - N8N_PROTOCOL=http
      - NODE_ENV=production

volumes:
  n8n_data:

Then:

# Clone the repository
git clone https://github.com/StudioExitt/n8n-nodes-comfyui-image-to-image.git /tmp/custom-nodes
cd /tmp/custom-nodes

# Install dependencies and build
npm install
npm run build

# Copy the built files to your custom nodes directory
mkdir -p ./custom-nodes/n8n-nodes-comfyui-image-to-image
cp -r dist/* ./custom-nodes/n8n-nodes-comfyui-image-to-image/

# Start the docker-compose setup
docker-compose up -d

Node Types

ComfyUI Image Transformer Node

This node allows you to transform and enhance images using ComfyUI's image-to-image capabilities.

ComfyUI Dual Image Transformer Node

This node processes two input images and generates one output image using ComfyUI workflows. Perfect for image blending, comparison, fusion, and other dual-image operations.

Settings

API URL: The URL of your ComfyUI instance (default: http://127.0.0.1:8188)
API Key: Optional API key if authentication is enabled
Workflow JSON: The ComfyUI workflow in JSON format for image transformation
Input Type: Choose between URL, Base64, or Binary input methods
Input Image: URL or base64 string of the input image (when using URL or Base64 input type)
Binary Property: Name of the binary property containing the image (when using Binary input type)
Timeout: Maximum time in minutes to wait for image transformation

Input

The node accepts an image input in three ways:

URL: Provide a direct URL to an image
Base64: Provide a base64-encoded image string
Binary: Use an image from a binary property in the workflow (e.g., from an HTTP Request node)

Outputs

The node outputs the transformed image:

In the binary.data property with proper MIME type and file information
fileName: Name of the transformed image file
data: Base64 encoded image data
fileType: The type of image file (e.g., 'image')
fileSize: Size of the image in KB
fileExtension: File extension (png, jpg, webp)
mimeType: MIME type of the image

ComfyUI Video Generator Node

This node generates videos from two input images using ComfyUI's video generation capabilities.

Settings

API URL: The URL of your ComfyUI instance (default: http://127.0.0.1:8188)
API Key: Optional API key if authentication is enabled
Workflow JSON: The ComfyUI workflow in JSON format with video generation nodes (AnimateDiff, SVD, etc.)
First Image Input Type: Choose between URL, Base64, or Binary for the first image
First Image: URL or base64 string of the first input image
First Image Binary Property: Name of the binary property containing the first image
Second Image Input Type: Choose between URL, Base64, or Binary for the second image
Second Image: URL or base64 string of the second input image
Second Image Binary Property: Name of the binary property containing the second image
First Image Node ID: Node ID in workflow for the first LoadImage node (default: "load_image_1")
Second Image Node ID: Node ID in workflow for the second LoadImage node (default: "load_image_2")
Video Frame Count: Number of frames to generate for the video (default: 16)
Video Frame Rate: Frame rate for the output video (default: 8 FPS)
Timeout: Maximum time in minutes to wait for video generation (default: 60)

Input

The node accepts two image inputs, each can be provided in three ways:

URL: Provide direct URLs to images
Base64: Provide base64-encoded image strings
Binary: Use images from binary properties in the workflow

Outputs

The node outputs the generated video:

In the binary.data property with proper MIME type and file information
fileName: Name of the generated video file
data: Base64 encoded video data
fileType: The type of file ('video' or 'image' for GIF)
fileSize: Size of the video in KB
fileExtension: File extension (mp4, webm, gif, mov)
mimeType: MIME type of the video
frameCount: Number of frames in the video
frameRate: Frame rate of the video
duration: Duration of the video in seconds

ComfyUI Audio Generator Node

This node generates audio from text prompts using ComfyUI's ACE Step Audio models.

Settings

API URL: The URL of your ComfyUI instance (default: http://127.0.0.1:8188)
API Key: Optional API key if authentication is enabled
Workflow JSON: The ComfyUI workflow in JSON format for text-to-audio generation
Text Prompt: Description of the audio you want to generate (e.g., "Jazz piano with smooth saxophone")
Audio Duration: Duration of the generated audio in seconds (default: 180)
Audio Quality: Bitrate quality of the output audio (64k, 128k, 192k, 256k, 320k)
Seed: Random seed for generation (-1 for random)
Steps: Number of sampling steps (default: 50)
CFG Scale: CFG scale for guidance strength (default: 5)
Timeout: Maximum time in minutes to wait for audio generation (default: 30)

Input

The node accepts a text prompt describing the desired audio output.

Outputs

The node outputs the generated audio:

In the binary.data property with proper MIME type and file information
fileName: Name of the generated audio file
data: Base64 encoded audio data
fileType: The type of file ('audio')
fileSize: Size of the audio in KB
fileExtension: File extension (mp3, wav, ogg, m4a)
mimeType: MIME type of the audio
prompt: The original text prompt used
duration: Duration setting used for generation
quality: Quality setting used for generation

Usage Examples

Using the ComfyUI Image Transformer Node

Create a workflow in ComfyUI for image transformation (e.g., img2img, inpainting, style transfer)
Export the workflow as JSON (API)
Add the ComfyUI Image Transformer node
Paste your workflow JSON
Select the appropriate Input Type:
- For URL: Enter the image URL
- For Base64: Provide a base64 string
- For Binary: Specify the binary property containing the image (default: "data")
Configure timeout as needed
Execute the workflow to generate a transformed image from your input image

Using the ComfyUI Dual Image Transformer Node

Create a workflow in ComfyUI for dual image processing (e.g., image blending, comparison, fusion)
Ensure your workflow contains at least 2 LoadImage nodes
Export the workflow as JSON (API)
Add the ComfyUI Dual Image Transformer node
Paste your workflow JSON
Configure both image inputs:
- Set input types for each image (URL, Base64, or Binary)
- Provide image sources for both images
- Specify binary property names if using Binary input type
Configure timeout as needed
Execute the workflow to generate a single output image from your two input images

Dual Image Workflow Requirements

Your ComfyUI workflow should include:

At least two LoadImage nodes (the first two will be used automatically)
Image processing nodes for blending, comparison, or fusion (e.g., ImageBlend, ControlNet, IPAdapter)
Output nodes like SaveImage or PreviewImage

Common use cases:

Image Blending: Combine two images with different blend modes
Style Transfer: Apply the style of one image to another
Face Swapping: Swap faces between two portrait images
Background Replacement: Replace the background of one image with another
Image Comparison: Create side-by-side or overlay comparisons

Using the ComfyUI Video Generator Node

Create a workflow in ComfyUI with video generation capabilities:
- Install AnimateDiff, SVD, or other video generation extensions
- Create a workflow with two LoadImage nodes (IDs: "load_image_1" and "load_image_2")
- Add video generation nodes (AnimateDiffSampler, VHS_VideoCombine, etc.)
- Export the workflow as JSON (API)
Add the ComfyUI Video Generator node
Paste your workflow JSON
Configure both image inputs:
- Set input types (URL, Base64, or Binary)
- Provide image sources
- Verify node IDs match your workflow
Set video parameters:
- Frame count (how many frames to generate)
- Frame rate (FPS)
- Timeout (video generation takes longer than images)
Execute the workflow to generate a video transitioning between your two input images

Example Video Workflow Requirements

Your ComfyUI workflow should include:

Two LoadImage nodes with specific IDs
Video generation nodes like:
- AnimateDiffLoader + AnimateDiffSampler
- SVD_img2vid_Conditioning for Stable Video Diffusion
- VHS_VideoCombine for video output
- LatentInterpolate for smooth transitions between images

See image_to_video_example.json for a complete workflow example.

Using the ComfyUI Audio Generator Node

Create or obtain a workflow in ComfyUI for audio generation:
- Install ACE Step Audio models and extensions
- Create a workflow with text encoding and audio generation nodes
- Include nodes like TextEncodeAceStepAudio, EmptyAceStepLatentAudio, VAEDecodeAudio, SaveAudioMP3
- Export the workflow as JSON (API)
Add the ComfyUI Audio Generator node
Paste your workflow JSON
Configure audio generation parameters:
- Enter your text prompt describing the desired audio
- Set audio duration (in seconds)
- Choose audio quality/bitrate
- Adjust sampling parameters (seed, steps, CFG scale)
- Set appropriate timeout (audio generation can take several minutes)
Execute the workflow to generate audio from your text prompt

Example Audio Workflow Requirements

Your ComfyUI workflow should include:

TextEncodeAceStepAudio node for text prompt encoding
EmptyAceStepLatentAudio node for latent audio space initialization
CheckpointLoaderSimple with ACE Step Audio model (e.g., ace_step_v1_3.5b.safetensors)
KSampler for the generation process
VAEDecodeAudio for decoding latent audio to actual audio
SaveAudioMP3 (or other audio save nodes) for output

See text_to_audio_example.json for a complete workflow example.

Error Handling

The node includes comprehensive error handling for:

API connection issues
Invalid workflow JSON
Execution failures
Timeout conditions
Input image validation

Development

# Install dependencies
npm install

# Build
npm run build

# Test
npm run test

# Lint
npm run lint

License

MIT

한국어 문서