Instructions to use Comfy-Org/MiniMax-H3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusion Single File
How to use Comfy-Org/MiniMax-H3 with Diffusion Single File:
# 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
Well, just in case anyone is wondering if can do i2v + ref, the answer is yes.
But it requires a little bit of custom conditioning. Here is the repo: https://github.com/kitsune123150/minimax-h3-hybrid-cond
First Frame
Reference Video
Output Video
The video shown earlier was made by fl2va model.
The workflow is already in the output. Just drag and drop it into Comfy.
Another Example:
| First Frame | Ref 1 | Ref 2 |
|---|---|---|
![]() |
![]() |
![]() |
Output:
:]
is it possible to add reference video in default workflow ?
The ComfyUI_Eclipse nodes do not install correctly, I already installed them manually and confyui still detects that they are not installed.
The ComfyUI_Eclipse nodes do not install correctly, I already installed them manually and confyui still detects that they are not installed.
You can skip it. it just ram cleanup.
is it possible to add reference video in default workflow ?
you can use it in R2V, but not in i2v.
What are you using to build such complex prompts? They almost look like code. Surely you aren't manually typing/formatting that.



