renmengwen /MuseDock

About MuseDock

MuseDock turns a topic, article, link or subtitle file into a short video: script, shots, voice, subtitles and render. It has two modes, HyperFrames motion video and hand-drawn whiteboard animation. It suits explainer clips you want to keep editing after the first render. You bring your own model API keys, the README is Chinese only, and the setup docs assume Windows.

Best for: People making explainer or whiteboard-style short videos who want a local, editable project instead of a one-shot MP4.

• Two modes: HyperFrames motion video and line-art whiteboard animation
• Starts from a topic, article, WeChat post, GitHub or Douyin link, image, or SRT file
• Builds HTML/CSS/GSAP shots you can edit and re-export in a project editor
• Whiteboard mode does 16:9, 9:16 and 4:3, with six paper templates plus a white handwritten style
• Transcribes public Douyin videos to TXT and SRT with local FunASR
• Logs every API reply, so a failed job resumes from the parts that worked

1. 1
   Install Node.js >=22 <23, Python 3.10+, Google Chrome and ffprobe.
2. 2
   In the project folder, run npm install (it also sets up the Python environment for whiteboard mode).
3. 3
   Start it with npm run dev, then open http://localhost:5173.
4. 4
   In Settings, add at least one analysis model with its API Key and Base URL. Add TTS, an image model and Pexels if you want voice, images or stock photos.
5. 5
   If the whiteboard media environment breaks, rerun npm run setup:whiteboard.