Skip to content
Core

Voice Control

Push-to-talk conversation with real speech synthesis.

Overview

Talk to Paw the way you'd talk to a colleague. Push-to-talk captures your voice, speech-to-text transcribes it, and Paw replies with synthesized speech that starts as soon as the first sentence is ready — not after the whole response is generated.

How it works

  1. 1

    Hold to talk

    Press and hold the push-to-talk control, or use the configured hotkey.

  2. 2

    Speak naturally

    Your speech streams to the speech-to-text provider as you talk.

  3. 3

    Paw responds

    The reply streams back and is spoken sentence-by-sentence, not held until it's fully written.

Use cases

  • • Dictating requests while your hands are on the keyboard for something else
  • • A more natural interaction than typing for quick questions

Limitations

  • • Voice quality depends on the configured provider — PawOS doesn't train its own voice models
  • • Background noise can affect transcription accuracy, as with any speech-to-text system