A lightweight cross-platform app that converts speech to text and processes it with AI. Perfect for crafting prompts for AI tools like Claude and ChatGPT - speak in your native language, get polished English output ready to use.
We strongly recommend using Homebrew for macOS. It handles installation, updates, and uninstallation automatically.
brew tap efeecllk/voice-prompt
brew install --cask voice-promptWorks on both Apple Silicon (M1/M2/M3) and Intel Macs automatically.
Why Homebrew?
- Automatic architecture detection (Apple Silicon vs Intel)
- Easy updates with
brew upgrade - Clean uninstall with
brew uninstall - No manual DMG mounting needed
Note: Direct DMG downloads are available in Releases but are not recommended. Use Homebrew instead.
Download from Releases:
Voice.Prompt_x.x.x_x64.msi- Windows Installer (recommended)Voice.Prompt_x.x.x_x64-setup.exe- NSIS Installer
Note: You may see a SmartScreen warning on first run. Click "More info" → "Run anyway".
macOS:
brew update && brew upgrade --cask voice-promptWindows: Download the latest version from Releases.
| Feature | Description |
|---|---|
| Menu Bar / System Tray | Lives in your menu bar (macOS) or system tray (Windows), always one click away |
| Voice Recording | One-click recording with visual feedback and waveform animation |
| Speech-to-Text | Powered by OpenAI Whisper API - industry-leading accuracy |
| Multi-Language | 20+ source languages including auto-detect |
| Global Shortcut | Trigger recording from any app (default: Cmd/Ctrl + Shift + Space) |
| Dark Mode | Follows your system theme automatically |
| Format | Description |
|---|---|
| Standard Translation | Clean, accurate English translation |
| Formal English | Professional business English |
| Casual English | Friendly, conversational tone |
| Ralph Wiggum Mode | Transforms simple requests into detailed Claude Code prompts |
| Custom Formats | Create your own AI processing templates |
| Feature | Description |
|---|---|
| History | Access your last 20 translations |
| Favorites | Star important outputs for quick access |
| My Prompts | Save outputs to your personal prompt library |
| One-Click Copy | Copy any text instantly to clipboard |
| Send to Terminal | Paste prompts directly into Ghostty, Warp, iTerm2, or Terminal.app |
Click the microphone button or use the global shortcut to start recording. The app shows:
- Recording duration timer
- Live audio waveform visualization
- Clear stop button
Recording stops automatically when you click stop, and processing begins immediately.
Choose how your speech is processed:
- Default Translation - Translates to natural English
- Formal Business - Professional tone for work contexts
- Casual Conversation - Relaxed, friendly output
- Ralph Wiggum Mode - Special mode that expands simple voice requests into detailed, well-structured prompts for Claude Code
Create your own processing templates:
- Go to Settings
- In "Create Your Own Output Format", click the microphone icon
- Describe what you want in your voice (e.g., "Convert my speech into git commit messages")
- The AI generates a custom format with name, description, and system prompt
- Optionally select an emoji icon for easy recognition
- Click Save
Your custom format appears in the Output Format dropdown and shows in the header when active.
Save and organize your favorite outputs:
- From Favorites: Click the document icon on any favorite to save it
- Access: Click the document icon in the header to view My Prompts
- Edit: Click any prompt to edit its name, description, or content
- Copy: Copy individual fields (name, description, content) with one click
- Delete: Remove prompts you no longer need
- History Tab: Shows your last 20 recordings with source and output
- Favorites Tab: Shows starred items for quick access
- Star Button: Click to add/remove from favorites
- Copy Button: Copy the output text instantly
- Save to My Prompts: Save favorites to your prompt library
Trigger Voice Prompt from any application:
| Shortcut | Platform |
|---|---|
Cmd + Shift + Space |
macOS (default) |
Ctrl + Shift + Space |
Windows (default) |
Change in Settings → Global Shortcut. Available options:
Cmd/Ctrl + Shift + SpaceCmd/Ctrl + Option/Alt + SpaceCmd/Ctrl + Shift + .Cmd/Ctrl + Option/Alt + V
Paste generated prompts directly into your terminal app without manual copy-paste:
- The app auto-detects running terminals (Ghostty, Warp, iTerm2, Terminal.app)
- Click the arrow button on any result to send it to your terminal
- Or enable Auto-paste in Settings to send automatically after generation
- Optionally enable Auto-submit to also press Enter after pasting
Supported terminals: Ghostty, Warp, iTerm2, Terminal.app
Setup: Grant Accessibility permission on first use (System Settings > Privacy & Security > Accessibility).
- System (default) - Follows your OS appearance
- Light - Always light mode
- Dark - Always dark mode
| Language | Code | Language | Code |
|---|---|---|---|
| Auto-detect | auto | Korean | ko |
| Turkish | tr | Chinese | zh |
| English | en | Arabic | ar |
| Spanish | es | Hindi | hi |
| French | fr | Dutch | nl |
| German | de | Polish | pl |
| Italian | it | Swedish | sv |
| Portuguese | pt | Danish | da |
| Russian | ru | Norwegian | no |
| Japanese | ja | Finnish | fi |
| Greek | el |
Your OpenAI API key is stored securely on your device.
- Get your API key from platform.openai.com
- Open Voice Prompt → Settings
- Paste your API key
- Click Save
| Service | Cost |
|---|---|
| Whisper (Speech-to-Text) | ~$0.006/minute |
| GPT-4.1-nano (Processing) | ~$0.0001/request |
Estimated: ~$0.05/day for 50 translations
- Node.js 20+
- pnpm 9+
- Rust stable
- macOS: Xcode Command Line Tools
- Windows: Visual Studio Build Tools with C++ workload
# Clone the repository
git clone https://github.com/efeecllk/voice-prompt.git
cd voice-prompt
# Install dependencies
pnpm install
# Run in development
pnpm tauri dev
# Build for production
pnpm tauri build| Layer | Technology |
|---|---|
| Framework | Tauri 2.0 |
| Frontend | React 18 + TypeScript |
| Styling | Tailwind CSS |
| State | Zustand |
| Speech-to-Text | OpenAI Whisper |
| Text Processing | OpenAI GPT-4.1-nano |
xattr -cr /Applications/Voice\ Prompt.appClick "More info" → "Run anyway". This appears because the app isn't code-signed.
- Check your API key at platform.openai.com
- Ensure you have billing enabled
- Try generating a new key
- Grant microphone permission when prompted
- Check System Settings → Privacy & Security → Microphone (macOS)
- Check Settings → Privacy → Microphone (Windows)
- Grant Accessibility permission: System Settings → Privacy & Security → Accessibility → Add Voice Prompt
- Restart the app after granting permission
- Check that the correct terminal is selected in Settings → Target Terminal
MIT License - see LICENSE for details.
Contributions are welcome! Please feel free to submit a Pull Request.
- Send to Terminal - Paste prompts directly into Ghostty, Warp, iTerm2, or Terminal.app
- Auto-paste - Automatically send generated prompts to your terminal
- Auto-submit - Optionally press Enter after pasting for hands-free workflow
- Smart Terminal Detection - Auto-detects running terminals, prioritizes dev terminals
- History Send - Send any history item directly to your terminal
- Windows Support - Full Windows 10+ support with native installer
- Intel Mac Support - Now works on both Apple Silicon and Intel Macs
- Custom Output Formats - Create your own AI processing templates with voice
- My Prompts Library - Save and organize your favorite outputs
- Favorites System - Star important translations for quick access
Made with Tauri