VoiceStudio

Project Overview

VoiceStudio (formerly OmniVoice-Studio) is a fully-local open-source AI voice studio offering voice cloning, voice design, video dubbing, dictation, transcription, and audiobook creation in 646 languages. It provides 16 TTS engines and 11 ASR engines, supports macOS, Windows, Linux, and Docker, with no account, API key, or subscription required.

Key Features

• Voice cloning and design
• Video dubbing
• Dictation and transcription
• Stories, audiobooks, and batch generation
• 646-language catalog (coverage depends on the selected engine)
• 16 TTS engines · 11 ASR engines
• Supports CUDA, Apple Silicon MPS/MLX, ROCm on Linux, CPU, and optional remote workers
• Desktop app, local REST/SSE/WebSocket API, OpenAI-compatible audio API, and MCP Server

Target Audience

• Content creators and video producers
• Developers who need speech synthesis and cloning capabilities
• Privacy-conscious users who want to run AI voice tools locally

Quick Start

1. Download the package for your platform from the latest release
2. macOS 13.3+ (Apple Silicon): DMG; Windows 10/11: MSI; Linux: AppImage (glibc 2.39+ required)
3. Docker is also available (CUDA, ROCm, CPU, and worker-only GPU profiles)
4. First launch creates a managed Python environment and downloads the default model; later launches reuse both
5. Launch the app, open Voice Cloning, and add a clean voice sample (3 seconds works; 5-15 seconds recommended)

Core Advantages

• Fully local — data never leaves your machine
• No account, API key, or subscription required
• Supports 646 languages
• Desktop app and multi-protocol API
• AGPL-3.0 licensed

System Requirements

• OS: macOS 13.3+ (Apple Silicon), Windows 10/11 x64, Linux x86_64 (glibc 2.39+)
• Additional Dependencies: Intel Macs cannot run the local Python backend; use a remote backend instead