Instructions to use Rootport/LongVideo-30s-Workflow-Wan2.2Remix-IV with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Wan2.2
How to use Rootport/LongVideo-30s-Workflow-Wan2.2Remix-IV with Wan2.2:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Wan2.2 Remix I2V β Long Video (~30s) Workflow
Image-to-Video workflow for generating ~30-second videos using Wan2.2 Remix v3.0 with a two-pass (LowβHigh) sampling approach.
Overview
- 7 segments Γ 81 frames each, chained sequentially (last frame β next segment's input)
- Two-pass KSampler: Low model (steps 0β10) β High model (steps 10β12)
- Output: 720Γ1280, 16fps, H.264 MP4
Required Models
| Type | File |
|---|---|
| UNET (Low) | wan22RemixT2VI2V_i2vLowV30.safetensors |
| UNET (High) | wan22RemixT2VI2V_i2vHighV30.safetensors |
| CLIP | nsfw_wan_umt5-xxl_fp8_scaled.safetensors |
| VAE | Wan2_1_VAE_bf16.safetensors |
Required Custom Nodes
- ComfyUI-VideoHelperSuite β
VHS_VideoCombine - ComfyUI-PainterVRAM β VRAM management
- rgthree-comfy β
Seed
Usage
- Load an image into the
LoadImagenode - Edit the positive prompt for each segment to describe scene progression
- Queue and wait β each segment generates sequentially
Each segment's prompt is independent, so you can tell a story that evolves across the 7 segments.
- Downloads last month
- -