Wan2.1

Why we included this project

Wan2.1 is a full video generation suite rather than a single model, covering text-to-video, image-to-video, video editing, and video-to-audio from one codebase. The 1.3B text-to-video variant runs on consumer hardware with about 8 GB of VRAM and can produce short 480P clips, so small studios and individual creators can use it without renting GPU clusters. It also renders legible Chinese and English text inside generated frames, a capability most video models lack and one that matters for ads and captions. Since it plugs into ComfyUI and Diffusers, teams can add it to existing pipelines instead of building a generation stack from scratch.