A powerful and user-friendly voice cloning tool that allows you to clone voices from audio samples and generate speech in multiple languages using state-of-the-art AI technology.

Features

🎯 Voice Cloning: Clone any voice from WAV audio samples
🌍 Multi-language Support: Generate speech in various languages
🎥 YouTube Integration: Download voice samples directly from YouTube videos
🔊 Audio Processing: Automatic silence removal and audio cleaning
🖥️ Cross-platform: Works on Windows, macOS, and Linux
🎛️ User-friendly CLI Interface: Easy-to-use menu system

Requirements

Python 3.10.16 (or lower, mandatory for TTS to be installed)
Internet connection for model download (first run only) and voice download (if needed)

Installation

NOTE: Skip this section if you are using uv (recommended)

pip install -r requirements.txt

Then, copy the .env.example file to .env:

cp .env.example .env

And adjust it accordingly. Anyway, you can override the configuration at runtime.

Usage

NOTE: If you are using uv, dependencies will be resolved in a .venv file at runtime

IMPORTANT: The included voice sample is noisy and short, so the result might be low quality. Use a better one for production. Sorry.

Using uv

uv run src/main.py

Normal python

python src/main.py

Overview

The application provides an interactive menu with the following options:

Start voice cloning with current settings
Select a target voice from available samples
Set a custom sentence to generate
Choose the target language
Download new voice samples from YouTube
Reset settings to default
Exit (duh)

Voice Sample Guidelines

Use clear, high-quality audio samples
Samples should be in WAV format
Ideal sample length: 10-30 seconds
Avoid background noise or music
Place voice samples in the data/ directory

Supported Languages

Use two-letter language codes (e.g., 'en' for English, 'fr' for French, 'es' for Spanish)