ComfyUI-Seedance-Nodes

Custom nodes for ComfyUI that enable video generation using ByteDance's Seedance model via Fal.ai.

Features

🎬 Seedance Image to Video (Fal.ai)

Generate high-quality videos from images using ByteDance's Seedance model.

Capabilities:

• Image-to-video generation
• Pro (1080p) and Lite (720p) models
• Multiple aspect ratios (auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16)
• Configurable duration (3-12 seconds)
• Camera control (fixed or dynamic)
• Seed-based reproducibility

📥 Download Video from URL

Automatically download generated videos to your ComfyUI output folder.

Features:

• Downloads video from URL
• Saves with timestamp
• Custom filename prefix
• Returns local file path

Installation

1. Clone this repository into your ComfyUI custom_nodes folder:

cd ComfyUI/custom_nodes
git clone https://github.com/YOUR_USERNAME/ComfyUI-Seedance-Nodes.git

2. Install dependencies:

cd ComfyUI-Seedance-Nodes
pip install -r requirements.txt

3. Restart ComfyUI

Requirements

• Python 3.8+
• ComfyUI
• fal-client
• Fal.ai API key

Getting Your Fal.ai API Key

1. Sign up at Fal.ai
2. Go to API Keys Dashboard
3. Click Create Key
4. Copy your API key

Usage

Basic Workflow

1. Load Image → Connect to Seedance Image to Video
2. Add your Fal.ai API key to the node
3. Configure parameters (prompt, resolution, duration, etc.)
4. Connect video_url output → Download Video from URL
5. Run the workflow
6. Video saves to ComfyUI/output/seedance_video_[timestamp].mp4

Example Workflow

Load Image 
    ↓
Seedance Image to Video (Fal.ai)
    - api_key: YOUR_FAL_API_KEY
    - prompt: high-end real-estate promo video, smooth camera motion
    - model_version: pro
    - resolution: 1080p
    - duration: 5
    ↓ (video_url)
Download Video from URL
    - filename_prefix: my_video
    ↓
Video saved: ComfyUI/output/my_video_20251022_120000.mp4

Pricing

Fal.ai Costs (approximate)

• Pro (1080p): ~$0.62 per 5-second video
• Lite (720p): Lower cost alternative
• Free tier available for testing

Node Parameters

Seedance Image to Video

| Parameter | Type | Default | Description | |-----------|------|---------|-------------| | image | IMAGE | - | Input image | | prompt | STRING | - | Text description for motion | | fal_api_key | STRING | - | Your Fal.ai API key | | model_version | CHOICE | pro | pro (1080p) or lite (720p) | | resolution | CHOICE | 1080p | 1080p, 720p, or 480p | | aspect_ratio | CHOICE | auto | Video aspect ratio | | duration | INT | 5 | Video duration (3-12 seconds) | | seed | INT | -1 | Random seed (-1 for random) | | camera_fixed | BOOLEAN | False | Lock camera position |

Output: video_url (STRING) - URL to generated video

Download Video from URL

| Parameter | Type | Default | Description | |-----------|------|---------|-------------| | video_url | STRING | - | URL of video to download | | filename_prefix | STRING | seedance_video | Prefix for saved file |

Output: saved_path (STRING) - Local path to saved video

Troubleshooting

API key is required error

• Make sure you've entered your Fal.ai API key in the node
• Get a key at https://fal.ai/dashboard/keys

fal_client not installed error

• Run: pip install fal-client

Video download fails

• Check your internet connection
• Verify the video URL is valid
• Ensure you have write permissions in the output folder

Tips for Best Results

1. Prompts: Be specific about desired motion and camera movement
2. Camera Fixed: Enable for locked camera position (good for product shots)
3. Duration: Start with 5 seconds and adjust as needed
4. Resolution: Use Pro (1080p) for final outputs, Lite for testing

Example Prompts

• high-end real-estate agency promo video, subtle camera motion, variations in lighting
• smooth dolly shot moving forward, cinematic, professional
• gentle pan from left to right, revealing the scene
• lock camera but add life and subtle motion to the image

License

MIT License - See LICENSE file for details

Credits

• Built for ComfyUI
• Uses Fal.ai API
• ByteDance Seedance model

Support

For issues, questions, or feature requests, please open an issue on GitHub.

Version

Current version: 1.0.0

Changelog

v1.0.0 (2025-10-22)

• Initial release
• Seedance Image to Video node
• Download Video from URL node
• Support for Pro and Lite models
• Configurable parameters