#33 · Primary category: Video & Animation

Wan2.2

aigc video-generation

Wan: Open and Advanced Large-Scale Video Generative Models

Project last updated:03/17/26

GitHub Stars

17.3K

Forks

2.2K

Contributors

13

License

Apache-2.0

Why we included this project

Wan2.2 is a serious open-weight choice for teams building text-to-video or image-to-video pipelines. The repo ships inference code and weights for a 5B hybrid model that generates 720P video at 24fps, using a Mixture-of-Experts design that expands capacity without inflating compute cost. It is built to run on a single 4090, so small studios and solo developers can actually use it without a GPU cluster. Alongside the core model there are specialized variants for character animation and audio-driven cinematic video, and Diffusers and ComfyUI integrations make it easy to slot into existing workflows. The training data emphasizes cinematic aesthetics with detailed labels for lighting, composition, contrast, and color tone, which shows up in the controllable output quality.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category