autoclip

Overview

AutoClip is an open-source AI video clipping and highlight generation tool. It imports long videos, uses AI to analyze subtitles, identifies highlight segments, and automatically cuts them into short videos and compilations. It is positioned for interviews, podcasts, courses, talking-head videos, and livestream replays, and is offered as a desktop application, a Docker web interface, and a CLI / MCP for interactive or batch use. Input can come from local video files, YouTube or Bilibili links, optionally accompanied by SRT subtitles; when no subtitles exist, local Whisper transcription can be used. The system generates an outline and topic timeline from subtitles, scores segments, and extracts titles. It automatically produces clips, allows previewing results, and lets users adjust compilation order. Export presets are provided for Douyin, Xiaohongshu, YouTube Shorts, and Bilibili, with support for burning in subtitles and title cards. Model support includes cloud services such as Qwen, OpenAI-compatible interfaces, and Gemini, as well as local models through Ollama or LM Studio. The interface supports Chinese, English, Japanese, Korean, Spanish, Portuguese, Russian, and French, switchable from the top bar or following the system, while source material and generated content keep their original language. AutoClip is free and open source under the MIT license.