Overview
Wan 2.1 is a suite of video generation models released by Alibaba in February 2025 under the permissive Apache 2.0 license. It includes a 14-billion-parameter model for high quality and a 1.3-billion-parameter model that can run on consumer GPUs, covering text-to-video, image-to-video and video editing tasks.
Highlights
- Text-to-video and image-to-video generation
- 1.3B model runs on consumer GPUs; 14B model for higher quality
- Apache 2.0 license allowing commercial use
- Support for Chinese and English text in videos
- Active open-source ecosystem of tools and fine-tunes
Typical uses
Short video clips for creative projects, marketing tests, research on video generation and local experimentation.
Things to consider
- Generating video is compute-intensive, especially with the 14B model.
- Clips are short; longer narratives require editing.
- Newer Wan versions have been released since.
Specifications are taken from the provider’s official announcement, documentation or model card. Benchmark results depend on evaluation settings; see how AI benchmarks work.