Dictate listens to your voice, transcribes it to text, cleans up grammar and punctuation, and pastes it anywhere.
Requires a Mac with Apple silicon (M1 or later) and macOS 13 or later. On first launch it downloads its speech models once, about 3 GB.
Notes, docs, Claude, the terminal, you name it. Fast, accurate, and done.
Settings
Remap the activation key, switch languages, pick your voice-to-text model, toggle grammar correction, toggle auto-copy/paste, and set how long the widget stays on screen.
Watch it turn any voice recording into clean, clear, professional text on any surface.
The speech model runs on your Mac. Your voice is never uploaded, so it works on a plane and it keeps working if this site disappears. Built in Claude Code as a companion to Capture and Narrate: same principle across all three, the model runs locally, nothing you say or select ever leaves the machine.
- Native macOS app
- Runs fully on-device
