Alibaba AI models

Wan 3.0 is Alibaba Tongyi Lab's all-in-one multimodal video generation model for longer, reference-heavy, production-oriented work. One endpoint covers text-to-video, first- and last-frame image-to-video, reference-driven generation (up to 10 images, 5 videos, and 5 audio clips), localized video editing, and temporal extension, with documents and public webpages accepted as additional inputs. It generates 2-30 second clips at 480P, 720P, or 1080P with native synchronized audio on by default. API access only; no open weights.

Wan 2.7 is Alibaba Tongyi Lab's flagship video generation model, offering text-to-video, image-to-video, reference-to-video, and instruction-based video editing through one endpoint. It supports first-and-last-frame control, up to 5 reference images, videos, and audio clips for consistent subject and voice identity, native synchronized audio (ambient sound, lip-synced dialogue, music), connected multi-shot generation, and clips of 2-15 seconds at up to 1080P.

Wan 2.6 is a state-of-the-art, open-source multimodal AI video generation model developed by Alibaba Cloud (released around December 2025). It is designed to create high-fidelity, 1080p cinematic videos up to 15 seconds long from text, images, or reference videos.