Voicebox

Overview

Voicebox is a local-first voice cloning studio designed as a free and open-source alternative to cloud-based services like ElevenLabs. It allows users to download voice models, clone voices from short audio samples, and generate speech entirely offline on their own machine. The app includes DAW-like multi-track editing tools for composing multi-voice projects and offers hardware acceleration with Metal on macOS and CUDA on Windows/Linux for fast local inference. No Python installation is required, ensuring a streamlined native experience.

Key features

• Local voice cloning
• Offline TTS generation
• DAW-like editor
• Multi-voice projects
• Model downloads
• Privacy-focused
• Metal acceleration (macOS)
• CUDA acceleration (Windows/Linux)
• Native performance
• No Python required
• Open-source alternative