vibe-tts

A desktop Text-to-Speech application powered by Kokoro 82M, providing high-quality voice synthesis through an intuitive PyQt6 interface.

Features

• User-Friendly GUI: Clean PyQt6 interface for easy text-to-speech conversion
• Kokoro 82M Integration: Lightweight yet powerful TTS model with 82 million parameters
• Multi-Language Support: Support for 9 languages including English, Spanish, French, Hindi, Italian, Japanese, Portuguese, and Chinese
• Multiple Voices: Wide variety of male and female voices for each language
• Customizable Speech: Adjustable speed (50% - 200%) and volume control
• Real-time Playback: Instant audio playback with stop functionality
• Responsive Design: Threaded architecture keeps the UI responsive during synthesis
• Local Processing: Runs entirely on your machine without requiring external servers

Prerequisites

For Running the Application

• Python 3.12
• PyTorch (CPU or GPU support)
• No external server required - runs locally!

Installation

1. Clone the repository:

git clone https://github.com/yourusername/vibe-tts.git
cd vibe-tts

2. Install dependencies using uv:

uv sync

Usage

1. Launch the application:

uv run python tts_app.py

2. In the application:
   • Enter or paste text in the text area
   • Select language from the dropdown
   • Choose a voice for the selected language
   • Adjust speed (50% - 200%) and volume as needed
   • Click "Speak" to synthesize and play audio
   • Use "Stop" to halt playback

Configuration

Supported Languages

• American English
• British English
• Spanish
• French
• Hindi
• Italian
• Japanese (requires uv add misaki[ja])
• Brazilian Portuguese
• Mandarin Chinese (requires uv add misaki[zh])

Voice Examples

• American English: af_heart, af_bella, am_adam, am_michael
• British English: bf_emma, bf_isabella, bm_george, bm_lewis
• Spanish: ef_sofia, em_carlos
• French: ff_camille, fm_pierre
• And many more for each language!

Project Structure

vibe-tts/
- tts_app.py              # Main application with GUI and Riva integration
- main.py                 # Application entry point
- pyproject.toml          # Project dependencies and metadata
- README.md               # This file
- README_RIVA_SETUP.md    # Detailed Riva server setup guide
- riva_quickstart_v2.19.0/  # Riva setup scripts and configuration

Development

Running Tests

uv run pytest

Code Formatting

uv run ruff format .
uv run ruff check . --fix

Type Checking

uv run pyright

Troubleshooting

Installation Issues

• If you encounter errors with PyTorch, install it manually based on your system:
  • CPU: uv add torch --index-url https://download.pytorch.org/whl/cpu
  • CUDA: uv add torch --index-url https://download.pytorch.org/whl/cu121

Audio Issues

• Verify system audio is working
• Check volume settings in both the app and system
• Ensure soundfile is properly installed: uv add soundfile

Language-Specific Issues

• For Japanese support: uv add misaki[ja]
• For Chinese support: uv add misaki[zh]
• Some languages require espeak-ng: sudo apt-get install espeak-ng (Linux/WSL)

Performance

• GPU acceleration is supported automatically if CUDA is available
• For Apple Silicon Macs: The app will use MPS acceleration automatically
• First run may be slower as models are downloaded and cached

Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feature-name
3. Make your changes following the guidelines in CLAUDE.md
4. Run tests and formatters
5. Commit your changes
6. Push to your fork and submit a pull request

License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgments

• Built with Kokoro 82M by Hexgrad
• UI powered by PyQt6