Seedance ComfyUI Workflow Complete Guide

What is ComfyUI for Seedance?

ComfyUI is a node-based interface for advanced AI workflows that gives you granular control over Seedance video generation. Unlike the standard web interface, ComfyUI allows you to:

• Chain multiple generation steps
• Apply custom preprocessing
• Fine-tune individual parameters
• Automate batch processing
• Create reusable workflow templates

Prerequisites

Before starting, make sure you have:

• ComfyUI installed (version 0.1.0 or higher)
• Seedance API access (coming soon)
• Basic understanding of node-based interfaces
• 8GB+ VRAM recommended for local processing

Installing Seedance ComfyUI Node

Method 1: Via ComfyUI Manager

1. Open ComfyUI Manager
2. Search for "Seedance"
3. Click "Install" on the Seedance node package
4. Restart ComfyUI

Method 2: Manual Installation

cd ComfyUI/custom_nodes
git clone https://github.com/seedance/comfyui-seedance
cd comfyui-seedance
pip install -r requirements.txt

Basic Seedance Workflow

Step 1: Text-to-Video Node Setup

Create a basic text-to-video workflow:

1. Add Seedance Text2Video node
2. Add Load Prompt node
3. Add Preview Video node
4. Connect nodes: Prompt → Text2Video → Preview

Example Prompt:

A cyberpunk dancer performing in a neon-lit alley,
rain falling, dramatic camera angles, vibrant colors,
cinematic lighting, 8k quality

Step 2: Parameter Configuration

Key parameters to adjust:

• Steps: 20-50 (higher = better quality, slower)
• CFG Scale: 7-12 (controls prompt adherence)
• Seed: -1 for random, set number for consistency
• Aspect Ratio: 16:9, 9:16, or 1:1
• Duration: 5-10 seconds

Step 3: Advanced Settings

For professional results:

• Motion Strength: 0.5-0.8 (controls movement intensity)
• Temporal Consistency: 0.7-0.9 (reduces flickering)
• Audio Sync: Enable for music video mode
• Upscale Factor: 1.5x or 2x for higher resolution

Advanced Workflows

Workflow 1: Character Consistency Pipeline

Create videos with the same character across multiple shots:

Character Reference → Image Encoder → Text2Video → Character Lock

1. Upload reference image of your character
2. Use Image Encoder to extract features
3. Apply Character Lock to maintain appearance
4. Generate multiple shots with consistent character

Workflow 2: Music Video Automation

Sync video generation with audio beats:

Audio File → Beat Detector → Scene Generator → Seedance → Compositor

1. Load your music track
2. Auto-detect beats and transitions
3. Generate scenes matching rhythm
4. Composite final video with audio

Workflow 3: Batch Style Transfer

Apply different styles to the same base video:

Base Video → Style Extractor → [Style 1, Style 2, Style 3] → Batch Seedance

Generate multiple style variations simultaneously.

Pro Tips for ComfyUI

Tip 1: Use ControlNet for Precise Control

Add ControlNet nodes to guide composition:

• Depth ControlNet: Control camera perspective
• Pose ControlNet: Guide character movements
• Edge ControlNet: Define scene boundaries

Tip 2: Optimize Processing Speed

• Use Queue System for batch processing
• Enable Low VRAM Mode if needed
• Cache Embeddings for repeated prompts
• Process at lower resolution first, then upscale

Tip 3: Create Reusable Templates

Save your workflows as templates:

1. Configure your ideal settings
2. Save as .json file
3. Share with team or community
4. Load template for consistent results

Common Issues and Solutions

Issue: Video Flickering

Solution:

• Increase Temporal Consistency to 0.85+
• Reduce Motion Strength to 0.6
• Use higher Step count (40+)
• Enable Frame Interpolation node

Issue: Character Changes Between Frames

Solution:

• Use Character Reference node
• Lock seed value for consistency
• Increase CFG Scale to 9-10
• Add Character Consistency node

Issue: Slow Processing

Solution:

• Reduce video duration to 5 seconds
• Lower resolution to 720p
• Decrease Step count to 25
• Use GPU acceleration if available

Issue: Audio Not Syncing

Solution:

• Enable Audio-Visual mode explicitly
• Check audio file format (WAV/MP3)
• Adjust sync offset in settings
• Use Beat Alignment node

Example Workflows (Download)

Workflow 1: Basic Text-to-Video

{
  "nodes": [
    {"id": 1, "type": "LoadPrompt", "data": {...}},
    {"id": 2, "type": "SeedanceText2Video", "data": {...}},
    {"id": 3, "type": "PreviewVideo", "data": {...}}
  ]
}

Download Full Workflow →

Workflow 2: Image-to-Video with Style

Download Workflow →

Workflow 3: Music Video Generator

Download Workflow →

API Integration (Coming Soon)

When API access is available, you'll be able to:

from comfyui_seedance import SeedanceAPI

api = SeedanceAPI(api_key="your-key")

result = api.generate_video(
    prompt="cyberpunk dancer in neon city",
    duration=5,
    aspect_ratio="16:9",
    style="cinematic"
)

video_url = result.get_video_url()

Community Workflows

Join our Discord to access:

• 100+ community-created workflows
• Weekly workflow challenges
• Expert tips and tutorials
• Workflow marketplace

Next Steps

Now that you've mastered ComfyUI basics:

1. Experiment with different node combinations
2. Share your workflows with the community
3. Join weekly workflow challenges
4. Build custom nodes for specific needs

Related Resources

• Seedance vs Sora vs Kling Comparison
• Character Consistency Guide
• Advanced Prompt Engineering

Ready to start creating? Try Seedance Free →