VoiceCloner

A desktop application for voice design and voice cloning powered by Qwen3-TTS.

Features

• Voice Design - Create custom voices from natural language descriptions
• Voice Cloning - Clone any voice from just 3 seconds of audio
• Built-in Recording - Record your voice directly in the app
• 10 Languages - English, Chinese, Japanese, Korean, German, French, Russian, Portuguese, Spanish, Italian
• Local Processing - All AI processing happens on your machine

Quick Start

Prerequisites

• macOS 11+, Windows 10+, or Linux
• Python 3.11+
• Rust (for building from source)
• NVIDIA GPU with 8GB+ VRAM (recommended) or CPU (slower)

Development Setup

# Clone the repository
git clone https://github.com/adibhanna/voicecloner
cd voicecloner

# Run setup script (installs Python deps, builds Rust)
./scripts/setup.sh

# Run the app
cargo run

Build Release App (macOS)

# Build the app bundle
./scripts/build-macos.sh

# The app will be at: target/bundle/VoiceCloner.app
open target/bundle/VoiceCloner.app

# Optionally create a DMG for distribution
./scripts/create-dmg.sh

How It Works

When you launch VoiceCloner:

1. The app automatically starts a local Python backend server
2. The backend loads Qwen3-TTS models (auto-downloaded on first use)
3. You can design voices, clone voices, or generate speech
4. All processing happens locally on your machine

System Requirements

Minimum (CPU mode)

• 16GB RAM (32GB recommended for 1.7B models)
• 15GB free disk space
• Microphone for voice cloning

Recommended (GPU mode)

• NVIDIA GPU with 16GB+ VRAM (for 1.7B models)
• NVIDIA GPU with 8GB+ VRAM (for 0.6B models)
• 32GB RAM
• 20GB free disk space

Project Structure

voicecloner/
├── src/                    # Rust frontend (iced GUI)
│   ├── main.rs
│   ├── app.rs              # Main application
│   ├── ui/                 # UI panels
│   ├── audio/              # Recording & playback
│   ├── backend/            # Backend client & process manager
│   └── state/              # App state & persistence
├── backend/                # Python backend (FastAPI + Qwen3-TTS)
│   ├── main.py             # API server
│   ├── tts_engine.py       # TTS model wrapper
│   └── requirements.txt
└── scripts/                # Build scripts
    ├── setup.sh
    ├── build-macos.sh
    └── create-dmg.sh

Acknowledgments

This project is powered by Qwen3-TTS from Qwen Team. Qwen3-TTS provides:

• High-quality text-to-speech synthesis
• Voice cloning from short audio samples
• Voice design from natural language descriptions
• Support for 10+ languages

License

MIT