Skip to main content

Voice

Dooze listens when you hold a key and speak. Everything is transcribed locally on your Mac using WhisperKit. Nothing is sent to a server for transcription.

Configuring your keys​

Go to Settings > Voice. You can assign your own keys for two voice actions independently:

  • Chat key: talks to Dooze
  • Dictate key: types into any app

Six options per slot: fn, Control, Control (right only), Option, Option (right only), and Space bar (hold). The same key cannot be assigned to both slots. Settings show live keycap previews so you can confirm your choices.

Talking to Dooze​

Press and hold your Chat key. Start talking. Release when you are done.

Dooze routes your words based on what it hears. A question or statement goes to Chat. A skill keyword like "summarize" or "polish" triggers that skill directly.

Dictation​

Hold your Dictate key and speak. Your words appear wherever your cursor is, emails, messages, documents, any text field.

This is passthrough dictation. It does not go to Dooze. It goes to your app.

Music detection​

If music is playing when you start speaking, Dooze pauses it and resumes when you finish.

Secure fields​

If the active app has a password field focused, voice input is suppressed. Nothing is transcribed.

Setup​

Go to Settings > Voice to enable or disable voice input, choose a transcription model (smaller is faster, larger is more accurate), and configure your activation keys.