HappyHorse 1.0

• 🎬 Text-to-video generation from Alibaba
• 🎯 Launches the HappyHorse 1.0 video family
• 🌍 Turns written prompts into motion clips
• 🏢 Part of Alibaba's broader generative lineup
• 🎨 Built for creative, prompt-driven video synthesis

Alibaba

Alibaba Group is a Chinese multinational technology company founded in 1999 and headquartered in Hangzhou, Zhejiang. Originally built around e-commerce and cloud computing, Alibaba has become one of the most prolific contributors to open-weight AI research,…

HappyHorse 1.0 is Alibaba's text-to-video model, generating short video clips directly from written prompts. Introduced in April 2026, it anchors a tightly clustered HappyHorse 1.0 release: companion endpoints arrived within days, including an image-to-video variant (HappyHorse 1.0), a reference-conditioned mode (HappyHorse 1.0 Reference), and a video editing mode (HappyHorse 1.0 Edit). Together they cover the common generative-video workflows under one family name.

Within Alibaba's wider catalogue, HappyHorse sits alongside the long-running Wan video line — including the recent Wan 2.7 family — as well as the Qwen text, image, and embedding models. As the text-to-video entry point, it complements rather than competes with those tools, focusing specifically on synthesizing animation from a textual description.

This is the current and newest model in the HappyHorse text-to-video line, so there is no in-family successor to consider. It is best suited to creators and developers who want to produce video clips from scratch using only a prompt, with the sibling reference and edit modes available when source imagery or existing footage is part of the pipeline.