Brands

Alibaba

Experimental visual generation models from Alibaba research teams.

Models

Video

Happy Horse

View details

Happy Horse is a general-purpose text-to-video and image-to-video model for clips from 3 to 15 seconds. In image-to-video mode, the uploaded image fixes the opening composition while the prompt directs motion; in text-to-video mode, the prompt and aspect ratio define the whole shot.

Alibaba HappyHorse 1.0

#Available

Happy Horse 1.0 generates 3–15 second video from text or animates one supplied first frame.

Wan 3.0 Video

View details

Wan 3.0 is Alibaba Cloud's all-in-one video model for text-led and reference-guided generation with native audio. It accepts first and last frames, up to ten images, five videos, and five audio references, or one document or public web page, with adaptive framing and 480p, 720p, or 1080p output.

Wan 3.0 Video

#Available

Generate cinematic videos from text, first and last frames, image, video, audio, document, or web references through Alibaba Cloud Model Studio.