VOX

An open-source, sophisticated multi-model AI audio generation platform

Integrating state-of-the-art voice conversion, SFX generation, and text-to-audio models into a seamless, high-fidelity experience.

---

Overview

VOX is a modular open-source AI audio platform that brings together state-of-the-art models for:

• Voice conversion & cloning
• Multilingual text-to-speech
• Text-to-audio & sound effects generation

Quick Start

One command sets up everything — environments, model weights, dependencies, and database:

chmod +x init.sh
./init.sh

Tech Stack

Frontend

• Next.js 15 (App Router)
• TypeScript
• Tailwind CSS
• Zustand
• Tanstack Query

Backend

• Node.js 20+
• Drizzle ORM
• p-queue

AI Models

• Seed-VC — Zero-shot voice conversion & cloning
• Make-An-Audio — Text-to-audio generation
• XTTS-v2 — High-quality multilingual TTS

Automation

• Bash orchestration
• Python-based environment & model manager

Project Structure

├── packages/
│   ├── app/          # Next.js frontend
│   └── server/       # Backend API & database
├── models/
│   ├── seed-vc/      # Voice conversion
│   ├── make-an-audio/# Audio generation
│   └── xtts-v2/      # Text-to-speech
├── data/             # Audio assets & outputs
└── init.sh           # One-command setup

System Requirements

• OS: macOS (MPS) or Linux (CUDA)
• Python: 3.10+
• Node.js: 20+
• GPU: Recommended (CPU supported with reduced performance)

License

MIT — free to use, modify, and distribute.