Push-to-talk voice dictation for macOS - Type with your voice, powered by local AI.
VoiceType is a lightweight menu bar app that lets you dictate text into any application using a simple push-to-talk hotkey. It uses WhisperKit for fully local, private speech-to-text transcription - no internet connection required, your voice never leaves your Mac.
- Push-to-Talk - Hold your hotkey, speak, release to transcribe and type
- 100% Local & Private - Uses WhisperKit for on-device transcription, no data sent to servers
- Works Everywhere - Types into any focused application (editors, browsers, chat apps, etc.)
- Multiple Languages - Auto-detect or choose from 99 supported languages
- Customizable Hotkey - Set your preferred key combination
- Multiple Models - Choose accuracy vs. speed (tiny, base, small, medium, large-v3)
- Menu Bar App - Lives in your menu bar, no dock icon clutter
- Launch at Login - Start automatically when you log in
- Download the latest
VoiceType-x.x.x.dmgfrom Releases - Open the DMG and drag VoiceType to your Applications folder
- Open VoiceType from Applications
- Grant the required permissions when prompted
The app is signed and notarized by Apple, so it will open without any security warnings.
Requirements:
- macOS 14.0 (Sonoma) or later
- Xcode 15+ or Swift 5.9+ toolchain
# Clone the repository
git clone https://github.com/twissmueller/voice-type.git
cd voice-type
# Build the app
./Scripts/build-app.sh
# Run the app
open .build/release/VoiceType.app
# Or create a DMG for distribution
./Scripts/create-dmg.sh- Launch VoiceType - The microphone icon appears in your menu bar
- Grant Permissions - Click the menu bar icon and grant all required permissions in Settings
- Wait for Model - The Whisper model downloads on first launch (~50-150MB depending on model)
- Start Dictating - Hold Option+Shift+Space (default), speak, then release
VoiceType needs three macOS permissions to function:
| Permission | Why It's Needed |
|---|---|
| Microphone | To record your voice for transcription |
| Accessibility | To type the transcribed text into applications |
| Input Monitoring | To detect the global hotkey in any app |
All permissions can be granted from the Settings window (click the menu bar icon → Settings).
Click the VoiceType icon in your menu bar, then click Settings to:
- Change the hotkey - Click the hotkey field and press your new combination
- Select a model - Larger models are more accurate but slower
- Choose a language - Or use auto-detect for any of 99 languages
- Enable launch at login - Start VoiceType automatically
| Model | Size | Speed | Accuracy | Best For |
|---|---|---|---|---|
| tiny | ~40MB | Fastest | Basic | Quick notes, simple commands |
| base | ~75MB | Fast | Good | General dictation |
| small | ~250MB | Medium | Better | Most users (recommended) |
| medium | ~750MB | Slow | Great | When accuracy matters |
| large-v3 | ~1.5GB | Slowest | Best | Professional transcription |
- Check that Input Monitoring permission is granted in System Settings → Privacy & Security
- Make sure VoiceType is enabled (green status in menu bar)
- Try a different hotkey combination if conflicts with other apps
- Check that Accessibility permission is granted
- Ensure the model has finished loading (no loading indicator in menu)
- Try speaking closer to your microphone
- macOS 14 (Sonoma) or later is required
- Try removing and re-adding permissions in System Settings
- Swift 5.9 & SwiftUI - Modern Apple development
- WhisperKit - On-device speech recognition
- AVAudioEngine - Low-latency audio capture
- CGEvent - Global hotkey detection and keystroke emulation
Contributions are welcome! Please feel free to submit a Pull Request.
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
If you find VoiceType useful, consider supporting its development:
This project is licensed under the MIT License - see the LICENSE file for details.
- WhisperKit by Argmax for the amazing on-device Whisper implementation
- OpenAI Whisper for the original speech recognition model
Made with ❤️ for the macOS community