Skip to content

Video Generation Workflow ​

Learn to generate videos with ComfyUI.

Supported Models ​

Video Models ​

ModelDescriptionFeatures
SVDStable Video DiffusionImage to video
AnimateDiffAnimation generationStyle animation

Model Selection ​

NeedRecommended
Realistic videoSVD
Style animationAnimateDiff

SVD (Stable Video Diffusion) ​

Installation ​

ComfyUI/models/checkpoints/
svd.safetensors           # Base version
svd_xt.safetensors        # XT version (higher quality)

Image to Video Workflow ​

┌─────────────┐    ┌─────────────┐    ┌─────────────┐
│ LoadImage   │ →  │ SVDVideoGen │ →  │ VideoCombine│
│ (first frame)│   │ eration     │    │             │
└─────────────┘    └─────────────┘    └─────────────┘

Parameters ​

ParameterDescriptionRecommended
framesVideo frames14-25
fpsFrame rate12-30
motion_bucket_idMotion intensity50-127

AnimateDiff ​

Installation ​

ComfyUI/models/checkpoints/
# Needs AnimateDiff-compatible model

ComfyUI/custom_nodes/
ComfyUI-AnimateDiff-Evolved/

AnimateDiff Workflow ​

┌─────────────┐    ┌─────────────┐    ┌─────────────┐
│ CheckpointL │ →  │ AnimateDiff │ →  │ VAEDecode   │
│ oader       │    │ Loader      │    │             │
└─────────────┘    └─────────────┘    └─────────────┘

Motion Modules ​

ModuleDescription
mm_sd_v15_v2.ckptStandard animation
mm_sd_v15_v3.ckptEnhanced

Video Save ​

VideoCombine Node ​

Output settings:
- format: mp4 / gif
- fps: 24-30
- codec: h264 / vp9

Common Formats ​

FormatUseRecommended
MP4UniversalRecommended
GIFSocial mediaLoop

Tips ​

Dynamic Description ​

DescriptionEffect
flowing hairFlowing hair
wind, breezeWind effect
dancing, movingDynamic action

Troubleshooting ​

Q: Video generation slow ​

A:

  • Reduce frames
  • Use smaller resolution
  • Use AnimateDiff Turbo

Q: Video shaky ​

A:

  • Lower motion_bucket_id
  • Use stable prompts

Next Steps ​