debpalash/VoiceStudio
★ 99.8k
Python
AGPL-3.0
Free, local voice cloning in 646 languages without cloud limits
A solo dev just open-sourced a free replacement for ElevenLabs.
Cloud TTS charges per character and locks voices in proprietary silos.
First, commercial voice services charge high monthly subscription fees and cap your character usage.
Cloud limits · Per-character fees and vendor lock-in
Clone any voice from a clean sample and dub into 646 languages.
Second, VoiceStudio clones realistic voices from a short audio clip and supports six hundred forty-six languages.
646 languages · Native multilingual dubbing pipeline
Switch between fourteen state-of-the-art TTS engines on your local GPU.
Third, you can switch between fourteen integrated speech engines like Kokoro and XTTS right on your own machine.
Local engines · Kokoro, XTTS, Chatterbox on device
Run the full desktop pipeline locally with simple CLI commands.
Fourth, install it with pip and synthesize or clone speech entirely offline with no API keys required.
Quick start · Zero API keys, offline execution
Open source on GitHub with over nineteen thousand stars.
It is completely open source under the AGPL license with over nineteen thousand stars on GitHub.
debpalash/VoiceStudio · AGPL-3.0 · 19.4k stars · Python