Build a state-of-the-art video generation model that produces cinematic-quality output, driving adoption of Alibaba's video creation tools.
HappyHorse 1.1 released June 22, 2026; topped Artificial Analysis video rankings in April 2026.
Updated 22 Jun 2026Checked 11 Oct0 updates this week
HappyHorse accepts text prompts and image inputs to generate video, handling text-to-video and image-to-video generation.
The model generates realistic video at 1080p resolution with smooth motion and coherent composition.
Synthesizes audio and syncs dialogue to character movements in seven languages including English, Mandarin, Japanese, and French.
Unlike video models that stitch multiple clips, HappyHorse generates each complete video in a single pass for coherent results.
| Spec | HappyHorse |
|---|---|
| Video resolution | 1080pR (reported) |
| Frame rate | 24 fpsR (reported) |
| Maximum video length | 3–15 secondsR (reported) |
| Audio synthesis | Included with dialogue lip-syncR (reported) |
| Single-pass generation | Complete video in one passR (reported) |
R reported by the company
Alibaba Qwen
Qwen open-weight models, Qwen app
Qwen is Alibaba Cloud's AI model family and one of the most used open-weight lines, with hundreds of thousands of spin-off models online. Sizes range from under 1 billion to over 1 trillion parameters.
