v

video-use

Developed by browser-use

video-use is an open-source video editing tool by browser-use, allowing users to edit videos by chatting with agents like Claude Code or OpenClaw. It processes videos efficiently using word-level audio transcripts and on-demand timeline visual composites instead of heavy frame decoding. Features include automatic filler-word and silence removal, auto color grading, audio popping elimination, custom subtitle burning, and parallel sub-agent animation generation, complete with a self-evaluation loop for quality assurance.

  • Agent-driven conversational video editing
  • Dual-layer multimodal analysis with low token usage
  • Automatic filler word and silence removal
  • Closed-loop rendering self-evaluation and correction
  • Parallel sub-agent animation overlay generation
webdesktop