Dictate with voice input

You are in the editor with Ranger open. Let Ranger interview you covers Interview me. Start voice input stays disabled until a section has started.

1. Click Interview me so the placeholder reads Ask for changes. Start voice input sits in the composer next to Upload context. It is a waveform control. Its name is Start voice input while idle.

1. Click Start voice input. If you already accepted Voice transcription, recording starts. If consent is still required, the Voice transcription dialog opens. The body says Dusted will transcribe your voice using AI, that the recording is used only to generate content on Dusted, and that handling follows the Privacy Policy at /privacy. Click Accept to start, or Decline, Close, or Escape to dismiss without saving consent.

2. After Accept, the control is named Stop voice input and pulses when the session is live. Speak toward the microphone. Live words appear as italic text under Write your reply, then move into the field. You can type while you speak.

3. Click Stop voice input when you finish. Leftover live words flush into the field. Click Send reply when the reply is ready. You can send while the session is still live. That click stops voice input and sends the field plus the live words as one message.

4. After you send, Start voice input and Upload context disable until the turn finishes. If the browser blocks the microphone, Microphone access needed appears under Write your reply. Allow the microphone, then click Start voice input again. Voice input unavailable appears when Dusted cannot start or keep the session.

Closing Close assistant does not revoke Voice transcription. Return to Meet Ranger or Let Ranger interview you if the composer is still locked.