🎤 Whisper Paste - Hinglish Voice Transcription

A lightweight Windows system tray application that converts speech to text using OpenAI's Whisper API. Perfect for code-switching between Hindi and English!

✨ Features

• 🔵 System Tray Application - Runs silently in background
• 🔑 Easy API Key Setup - Set your API key directly from the tray menu
• 🎯 Two Output Modes:
  • Hinglish (Roman): Preserves Hindi words in Roman script
  • English: Translates everything to English
• ⌨️ Global Hotkey - Ctrl+Space to start/stop recording
• 📋 Auto-Paste - Transcribed text automatically pastes at cursor
• 🔴 Visual Feedback - Icon changes color when recording
• ⏱️ Long Recordings - Up to 5 minutes per session
• 💾 Settings Persistence - Remembers your API key and preferences

🚀 Quick Start

Prerequisites

• Windows 10/11
• Python 3.8+
• OpenAI API Key (Get one here)

Installation

1. Clone the repository

   git clone https://github.com/mvijay24/whisper-paste.git
   cd whisper-paste

2. Install dependencies

   setup.bat

3. Run the application

   start_silent.vbs

   Or simply double-click start_silent.vbs for completely silent startup!

4. Set your API key

   • Right-click the tray icon
   • Select "🔑 Set API Key..."
   • Enter your OpenAI API key
   • Click Save

📖 Usage

1. Look for the mic icon in your system tray (near clock)
2. Right-click the icon to access settings:
   • Set/Update API Key
   • Choose output mode (Hinglish or English)
3. Press Ctrl+Space to start recording
4. Speak in Hindi, English, or Hinglish
5. Press Ctrl+Space again to stop
6. Text automatically pastes at your cursor position!

Menu Options

• 🔑 Set API Key... - Add or update your OpenAI API key
• API Status - Shows connection status (✓ Connected or ⚠️ No API Key)
• 📝 Output Mode - Choose between Hinglish (Roman) or English
• 🐛 Enable and Display Debug Panel - Shows real-time execution logs
• 📊 Session Log Summary - Shows detailed report after each recording
• Exit - Properly closes the application

Examples

Hinglish Mode:

• You say: "Bhai ye file jaldi bhej de"
• Output: bhai ye file jaldi bhej de

English Mode:

• You say: "Bhai ye file jaldi bhej de"
• Output: brother send this file quickly

🛠️ Building Executable

To create a standalone .exe file:

build.bat

The executable will be created in the dist folder.

📁 Project Structure

whisper-paste/
├── whisper_tray.pyw      # Main application (no console window)
├── start_silent.vbs      # Silent launcher
├── start.bat            # Standard launcher
├── restart.bat          # Kill old instances & restart
├── setup.bat            # Install dependencies
├── build.bat            # Build executable
├── icon.ico             # Application icon
├── settings.json        # Saved settings (auto-created)
└── README.md            # This file

⚙️ Configuration

Settings are automatically saved in settings.json:

{
  "output_mode": "hinglish",
  "api_key": "sk-..."
}

🔧 Troubleshooting

Can't see the tray icon?

• Click "Show hidden icons" arrow in system tray
• Use restart.bat to kill old instances

API Key issues?

• Ensure you have a valid OpenAI API key
• Check your API usage limits at OpenAI Dashboard

No audio recorded?

• Check microphone permissions in Windows Settings
• Ensure default microphone is set correctly

Text not pasting or cursor errors?

• Close clipboard managers like BeefText, Ditto, or ClipboardFusion - they interfere with paste functionality
• Disable any text expander software temporarily
• If you see "[WinError 1402] Invalid cursor handle", it's likely due to clipboard manager interference

💰 Cost

• Uses OpenAI's Whisper API
• Approximately $0.006 per minute of audio
• See OpenAI Pricing

💡 Why "Toast"?

The small popup notification that shows transcribed text is called a "toast" because:

• It "pops up" like bread from a toaster
• It appears briefly and then disappears
• Common UI term from Android/Windows for temporary notifications
• Shows at the corner of the screen without interrupting workflow

🤝 Contributing

Feel free to open issues or submit pull requests!

📜 License

MIT License - feel free to use this in your projects!

🙏 Acknowledgments

• OpenAI for the amazing Whisper API
• The Python community for excellent libraries
• Special thanks to the Hinglish-speaking community!

---

Made with ❤️ for the Hinglish-speaking developers!