Voice Input

Instead of typing, you can speak. Voice input records your microphone, transcribes the audio, and drops the text into the input — where you can edit it before sending.

---

Starting a recording

1. Click the microphone button in a chat or text input — it appears when the input is empty

2. Or press Ctrl+M from the input

3. macOS asks for microphone permission the first time — click Allow

A recording overlay appears with:

• A recording timer
• A live waveform of your voice
• Cancel (Esc) — discard the recording
• Done (Enter) — finish and transcribe

---

Finishing and editing

1. Press Enter or click Done (Enter) when you finish speaking

2. A Transcribing… state appears while the audio is converted to text

3. The transcribed text lands in the input — edit it like any typed text

4. Press Enter to send

---

Choosing a microphone

If you have multiple input devices:

1. Open Settings → System Preferences

2. Under Audio Settings, pick your preferred microphone

See System Preferences for the rest of that page.

Transcription runs via Groq Whisper in the cloud, so voice input requires an internet connection. If a recording fails to transcribe, check your connectivity and see Troubleshooting.

---

Use cases

• Long prompts, fast: dictate a multi-paragraph brief into a Chat Node in the time it would take to type the first sentence.
• Thinking out loud: ramble an idea into a Text Node, then ask a connected chat to structure it.
• Hands-busy capture: press Ctrl+M and describe what you're looking at — a whiteboard, a book page — without switching to the keyboard.
• Accessibility: a full alternative to typing for anyone who finds sustained keyboard input difficult.