Skip to main content
2026/8/10 12:39 (UTC+8) · Docs Update · ByteDance 📚 First video-model Agent Skill docs are live: Seedance 2.0 Following the image API Skills, the docs center now ships a ready-to-use Agent Skill for Seedance 2.0 — put two files into Codex, Claude Code, OpenClaw, or any coding agent and generate videos with a single sentence. The zero-dependency script covers text-to-video, first-frame / first-and-last-frame image-to-video, and reference-image-to-video, switching between mini / fast / std via --model. It defaults to the cheapest mini with synced audio; video generation is an async task, and the script polls and downloads the finished clip locally. See Seedance 2.0 Agent Skill.
← Back to Live Updates · 📚 Monthly Archive