Your Voice, Transcribed Instantly - 100% Private, No Internet Required
Transform your voice into text anywhere on Windows with a simple keyboard shortcut. Whether you're writing emails, taking notes, or coding - Whisper4Windows puts professional speech-to-text at your fingertips, completely free and private.
Speak naturally, watch your words appear instantly in any application
Your voice never leaves your computer. No cloud, no API keys, no subscriptions. Everything runs locally using OpenAI's powerful Whisper model.
With GPU acceleration, transcribe 30 seconds of speech in 0.5-2 seconds. No waiting, no buffering - just instant results. CPU fallback available for all systems.
Press F9 from any app - Word, Chrome, Slack, VSCode - speak your thoughts, and watch text appear exactly where you need it.
Supports 99 languages with automatic detection. Switch between languages effortlessly or lock to your preferred ones for better accuracy. Language detection works automatically or you can specify a language for better performance.
- Press F9 - Minimal recording window appears at top of your screen
- Speak naturally - Watch the live visualizer respond to your voice
- Press F9 again - Text appears instantly where your cursor is
Beautiful wave animation shows processing in real-time - all happening locally on your device
Customize Whisper4Windows to work exactly how you want:
Fine-tune your keyboard shortcuts, model quality, and processing device
Choose your preferred languages, enable GPU acceleration, and control clipboard behavior
Select your microphone, adjust sound effects, and customize the experience
- 🎯 Model Quality - Balance speed vs accuracy (tiny, base, small, medium, large-v3)
- ⚙️ Processing Device - Auto-detect or force GPU/CPU with automatic fallback
- 🌍 Languages - Auto-detect all 99 languages or specify a language for better performance
- 🎤 Microphone - Select from all available input devices with device index selection
- ⌨️ Keyboard Shortcuts - Customize toggle (default: F9) and cancel (default: Escape) shortcuts
- 📋 Clipboard - Keep text in clipboard or auto-restore previous content
- 🎨 Themes - Light, dark, or system-matched appearance
- Download - Get
Whisper4Windows_0.1.0_x64_en-US.msi(~660MB) - Install - Run the installer
- Launch - Start from Start Menu or Desktop
- First Run - Whisper model downloads automatically (~500MB, one-time)
- Start Dictating! - Press F9 anywhere and speak
✅ CUDA libraries bundled! GPU acceleration works out-of-the-box if you have an NVIDIA GPU.
- Clone - Clone this repo
- Run - Double-click
START_APP.bat - Wait - First launch downloads the Whisper model (~500MB, one-time only)
- Start Dictating! - Press F9 anywhere and speak
That's it! No accounts, no configuration wizards, no complicated setup.
Open Notepad (or any text app), press F9, and say:
"Hello world, this is Whisper for Windows transcribing my voice in real-time, completely offline and private."
Watch it appear instantly! ✨
- 📝 Write Emails - Dictate emails 3x faster than typing
- 📄 Take Notes - Capture meeting notes without looking away
- 💬 Chat Messages - Respond quickly in Slack, Teams, Discord
- 📚 Write Documents - Draft reports and documentation hands-free
- 💻 Code Comments - Dictate code comments and documentation
- 🌐 Browse Faster - Search Google by voice
- ♿ Accessibility - Perfect for users who prefer or need voice input
Minimum:
- Windows 10/11
- 8GB RAM
- Microphone (built-in or external)
Recommended for GPU Acceleration:
- NVIDIA GPU (GTX 1060 or better)
- 16GB RAM
- ✅ No CUDA installation required! CUDA libraries are bundled with the MSI installer
Performance:
- GPU Mode: 30 seconds of speech → 0.5-2 seconds transcription ⚡
- CPU Mode: 30 seconds of speech → 5-10 seconds transcription
- Model Sizes: tiny (fastest) → base → small (recommended) → medium → large-v3 (most accurate)
📖 Installation guide → | Build your own MSI →
Q: Is this really free? A: Yes! 100% free and open-source. No subscriptions, no API keys, no hidden costs.
Q: Do I need an internet connection? A: Only for the first-time model download (~500MB). After that, everything works completely offline.
Q: Do I need to install CUDA for GPU acceleration? A: No! The MSI installer includes all necessary CUDA libraries. Just install and it works if you have an NVIDIA GPU.
Q: Which languages are supported? A: All 99 languages that Whisper supports - from Afrikaans to Zulu, including English, Spanish, French, German, Chinese, Japanese, Arabic, and many more.
Q: How accurate is it? A: Very! Using the same AI model that powers many commercial services. Accuracy improves with better quality models (Medium/Large) and GPU acceleration.
Q: Will this slow down my computer? A: No. The app only uses resources when actively recording and transcribing. It sits quietly in the system tray otherwise.
Q: Can I use it for work? A: Absolutely! Perfect for emails, documentation, meeting notes, and more. Since everything is local, it's suitable for confidential work.
Backend not starting?
cd backend
venv\Scripts\activate
pip install -r requirements.txt
python main.pyGPU not working?
- MSI installer: GPU should work automatically - check Device Manager for NVIDIA GPU
- From source: Run
TEST_GPU.batto diagnose, see INSTALLATION.md - App automatically falls back to CPU if GPU fails
Hotkey not working?
- Check system tray - app must be running
- Try restarting the app
- Some apps may block global hotkeys
📖 Full troubleshooting guide →
- INSTALLATION.md - Detailed setup guide
- BUILD.md - Building MSI installer for distribution
- TECHNICAL.md - Technical documentation for developers
- DEVELOPMENT_MODE.md - Development workflow guide
Found a bug? Have a feature request? Contributions are welcome!
- Fork the repository
- Create your feature branch
- Submit a pull request
MIT License - Free to use and modify for personal or commercial use.
Built with incredible open-source technology:
- Whisper by OpenAI
- faster-whisper by Systran
- Tauri framework
- FastAPI framework
Inspired by Superwhisper for Mac.
Ready to transform your workflow? Download Whisper4Windows now and experience the future of voice input on Windows. 🎤