--model wan / happyhorse on one shared Wan&HappyHorse group token; local images upload directly as base64 in our tests, no image host needed. See the Wan2.7 / HappyHorse Video Agent Skill.
← Back to Live Updates · 📚 Monthly Archive
Wan2.7 and HappyHorse Video Agent Skill Docs Released
Following Seedance 2.0, the docs center adds a shared Agent Skill for the Alibaba video models Wan2.7 and HappyHorse: one zero-dependency script covers text-to-video, image-to-video, reference-to-video, and video editing, switching series via —model, with local images uploaded directly.
2026/8/10 13:29 (UTC+8) · Docs Update · Alibaba
📚 Video Agent Skills, round two: Wan2.7 and HappyHorse
Following Seedance 2.0, the docs center ships a shared Agent Skill for the two Alibaba video series — drop two files into Codex, Claude Code, or any coding agent and generate clips with a single sentence. One zero-dependency script covers all four modes — text-to-video, first-frame image-to-video, reference-to-video, and video editing — picking the right model ID from the assets you pass, switching series via
← Back to Live Updates · 📚 Monthly Archive
← Back to Live Updates · 📚 Monthly Archive
此页面对您有帮助吗？
⌘I