https://github.com/bebsworthy/voicetype
Privacy-first dictation for macOS using local AI models
https://github.com/bebsworthy/voicetype
accessibility coreml dictation macos offline privacy speech-to-text swift swiftui whisper
Last synced: 3 months ago
JSON representation
Privacy-first dictation for macOS using local AI models
- Host: GitHub
- URL: https://github.com/bebsworthy/voicetype
- Owner: bebsworthy
- License: mit
- Created: 2025-07-06T20:12:01.000Z (3 months ago)
- Default Branch: main
- Last Pushed: 2025-07-06T23:18:12.000Z (3 months ago)
- Last Synced: 2025-07-07T00:28:13.567Z (3 months ago)
- Topics: accessibility, coreml, dictation, macos, offline, privacy, speech-to-text, swift, swiftui, whisper
- Language: Swift
- Homepage: https://github.com/bebsworthy/voicetype
- Size: 310 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# VoiceType
**Privacy-first dictation for macOS**
[](https://opensource.org/licenses/MIT)
[](https://www.apple.com/macos/)
[](https://swift.org)
[](https://github.com/yourusername/voicetype/actions)
## 🎯 What is VoiceType?
VoiceType is an open-source, privacy-first dictation tool for macOS that converts speech to text using local AI models. Unlike cloud-based solutions, VoiceType processes everything on your device, ensuring your voice never leaves your computer.
### ✨ Key Features
- **🔒 100% Privacy**: All processing happens on-device. No cloud, no data collection, no internet required
- **🚀 Fast & Accurate**: Real-time transcription with <5 second latency using OpenAI Whisper models
- **🌍 30+ Languages**: Built-in support for multiple languages with auto-detection
- **⌨️ Universal Compatibility**: Works with any macOS application that accepts text input
- **🎛️ Flexible Models**: Choose between speed and accuracy with multiple model sizes
- **🔌 Extensible**: Plugin system for custom audio processors and text injectors
- **📖 Open Source**: MIT licensed, community-driven development
## 🚀 Quick Start
### Download
Download the latest release from the [Releases](https://github.com/yourusername/voicetype/releases) page.
### First Launch
1. **Open VoiceType** - Look for the microphone icon in your menu bar
2. **Grant Permissions** - Allow microphone access when prompted
3. **Choose Your Model** - Select Fast (default) for quick results or Accurate for better quality
4. **Set Your Hotkey** - Default is `Ctrl+Shift+V`
5. **Start Dictating** - Press your hotkey in any app and start speaking!
## 📋 System Requirements
- macOS 12.0 (Monterey) or later
- 8GB RAM minimum (16GB recommended for larger models)
- Apple Silicon (M1/M2/M3) or Intel processor
- ~200MB disk space (plus model downloads)
## 🎮 How to Use
1. **Position your cursor** where you want to insert text
2. **Press your hotkey** (default: `Ctrl+Shift+V`)
3. **Speak clearly** for up to 5 seconds
4. **Watch your words appear** - VoiceType automatically inserts the text
### Pro Tips
- Speak naturally at a normal pace
- Minimize background noise for best results
- Use the Accurate model for technical terms
- Customize your hotkey in Settings
## 🛠️ Building from Source
### Prerequisites
- Xcode 15.0 or later
- macOS 13.0+ (for development)
- Apple Developer account (for code signing)
### Build Instructions
```bash
# Clone the repository
git clone https://github.com/yourusername/voicetype.git
cd voicetype/VoiceType
# Setup development environment
./Scripts/setup.sh
# Build the app
./Scripts/build.sh
# Run tests
./Scripts/test.sh
# Create release build
./Scripts/release.sh
```
### Development
```bash
# Open in Xcode
open VoiceType.xcodeproj
# Or use Swift Package Manager
swift build
swift test
```
## 🔧 Configuration
VoiceType can be customized through its settings panel or by editing the configuration file:
`~/Library/Application Support/VoiceType/config.json`
### Available Settings
- **Hotkey**: Customize your recording trigger
- **Model Selection**: Choose between Tiny (fast), Base (balanced), or Small (accurate)
- **Language**: Select from 30+ languages or use auto-detection
- **Audio Device**: Choose your preferred microphone
## 🤝 Contributing
We welcome contributions! Please see our [Contributing Guide](Documentation/DeveloperGuide/Contributing.md) for details.
### Areas for Contribution
- 🔌 **App-specific text injectors** - Add support for more applications
- 🎤 **Audio preprocessors** - Improve noise reduction and audio quality
- 🌐 **Translations** - Help translate the UI to more languages
- 📚 **Documentation** - Improve guides and tutorials
- 🐛 **Bug fixes** - Help us squash bugs
## 🏗️ Architecture
VoiceType uses a modular, protocol-first architecture:
```
┌─────────────┐ ┌──────────────┐ ┌──────────────┐
│ Menu Bar │────▶│ Coordinator │────▶│Audio Processor│
└─────────────┘ └──────────────┘ └──────────────┘
│ │
▼ ▼
┌──────────────┐ ┌──────────────┐
│ Transcriber │────▶│Text Injector │
└──────────────┘ └──────────────┘
```
See our [Architecture Guide](Documentation/DeveloperGuide/Architecture.md) for details.
## 🐛 Troubleshooting
### Common Issues
**VoiceType doesn't appear in menu bar**
- Check if the app is running in Activity Monitor
- Try launching from Applications folder
**Hotkey doesn't work**
- Grant Input Monitoring permission in System Settings → Privacy & Security
- Check for conflicts with other apps
**No text appears after speaking**
- Verify microphone permission is granted
- Check audio input levels in Settings
- Try the clipboard fallback mode
See our [Troubleshooting Guide](Documentation/UserGuide/Troubleshooting.md) for more solutions.
## 📜 License
VoiceType is released under the MIT License. See [LICENSE](LICENSE) file for details.
## 🙏 Acknowledgments
- [OpenAI Whisper](https://github.com/openai/whisper) for the amazing speech recognition models
- [Apple CoreML](https://developer.apple.com/machine-learning/core-ml/) for on-device inference
- The Swift and macOS developer communities
## 🔗 Links
- [Documentation](Documentation/UserGuide/README.md)
- [Report Issues](https://github.com/yourusername/voicetype/issues)
- [Discussions](https://github.com/yourusername/voicetype/discussions)
- [Changelog](CHANGELOG.md)
---
Made with ❤️ for privacy-conscious users everywhere