Instructions to use Kijai/WanVideo_comfy with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusion Single File
How to use Kijai/WanVideo_comfy 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
Licensing of Lightx2v/lightx2v_I2V_14B_480p_cfg_step_distill_rank256_bf16.safetensors — confirming upstream equivalence
Hi Kijai (Jukka),
Thank you for maintaining this repository. It is genuinely load-bearing for a lot of us.
Context. I am publishing a Japanese book on image and video generation (ships 2026-09-19): https://j.aicu.ai/SG26
Chapter 5 walks readers through Wan2.2 I2V in ComfyUI on Colab, and it names your repository and this exact file in print:
Lightx2v/lightx2v_I2V_14B_480p_cfg_step_distill_rank256_bf16.safetensors
What I would like to confirm. This repository has no license: field in its cardData and no LICENSE file. Your README states that the models originate from Wan-AI and lightx2v, and both of those publish under Apache-2.0.
Is it correct to treat the files under Lightx2v/ as Apache-2.0, following upstream?
A one-word yes or no is all I need. I am not asking you to add a LICENSE file.
If I do not hear back by 2026-09-03, I will present the file as-is: attributed to you by name and repository, with no license claim of my own, and with a note that the redistributor has not declared one. I would rather print your answer than my assumption, which is why I am asking.
Either way — thank you for the work. It reaches further than you probably see.
— Akihiko SHIRAI, Ph.D. / AICU Japan K.K.