Repository navigation
Utter 0.3.0 (beta)
Utter 0.3.0 is the first beta. Free, open-source voice dictation for macOS: hold a key, speak, let go, and clean, well-formatted text appears in whatever app you are typing in.
Download: Utter-apple-silicon.dmg below. Apple Silicon Macs (M1 or later), macOS 12 or later.
What it does
- Fast. Speech is transcribed on your Mac while you are still talking, and tidied a paragraph at a time, so even a long dictation goes in a moment after you let go.
- Private by default. Audio never leaves your Mac. No account, no analytics.
- Writes for where you are. Email in Mail or Gmail, casual in Slack, a clear prompt in Claude or ChatGPT, exact identifiers in code editors.
- Your model, or none. Tidy with Groq, Anthropic, OpenAI, Gemini, OpenRouter or a local model through Ollama, with your own key.
- One key. Hold fn (globe) to talk; double-tap for hands-free. Or use any key combination.
- Rewrite and edit by voice. A second hotkey rewrites what you said in a style you pick, or, with text selected, changes it the way you ask.
- Your words and your history. A personal dictionary for names and jargon, and your recent dictations kept on your Mac to copy again.
- Tidy text. Short paragraphs, lists where you name several things, no em dashes.
Install
- Open the DMG and drag Utter into Applications.
- Open Utter. It is signed with a Developer ID and notarised by Apple, so it opens like any other app.
- Follow the welcome screens: allow the microphone and Accessibility, and let the speech model download.
Permissions now survive updates.
It's a beta
Tested day to day on one Mac (M1 Pro). Please report anything odd through the issue forms.
SHA-256: 0d1a471c0150e1a9a43bbb4c69d8c152952917b768c8b480a4ad8d026ae60663