# yap

A macOS CLI for on-device speech transcription using Apple's Speech framework, producing SRT/VTT/JSON captions from files, transcribing live system audio and microphone input, and exposing an MCP server for agents.

yap transcribes audio and video files on-device using Speech.framework on macOS 26, and can output SRT captions, WebVTT subtitles, or JSON with word-level timestamps. It composes with other command-line tools, for example piping a video download directly into yap, or piping a transcript into a local LLM for summarization or translation.

Beyond file transcription, yap listen transcribes system audio live, yap dictate transcribes microphone input, and yap listen-and-dictate transcribes both simultaneously with custom speaker labels, useful for meeting transcripts; each requires the relevant Screen Recording or Microphone permission for the terminal app.

It also ships an MCP server exposing a transcribe tool so any MCP-compatible agent, such as Claude Code, can transcribe audio and video files, and installs via Homebrew or Mint.

Website: https://github.com/finnvoor/yap