All models
voice-cloning

GPT-SoVITS

MIT

The community's favorite few-shot voice-cloning toolkit

GPT-SoVITS has one of the largest communities in open voice AI. It does zero-shot cloning from a few seconds of audio, but its signature trick is few-shot training: with just one minute of a voice it produces a dramatically more faithful clone. That workflow made it the de facto standard for hobbyist voice cloning, especially across the Chinese and Japanese communities.

Voices

1 voice
Neutral
Reference-baseddefault
—
NeutralEmotionalNumbers & Dates

The scripts

Neutral
“The quick brown fox jumps over the lazy dog near the riverbank.”
Emotional
“I can't believe you actually did it. This is incredible!”
Numbers & Dates
“On March 14th, 2025, the team raised $4.2 million at a 38% margin.”

Install

git clone https://github.com/RVC-Boss/GPT-SoVITS

GPT-SoVITS FAQ

Do I need a GPU to run GPT-SoVITS?

For practical speeds, yes — plan on roughly 4 GB of VRAM. If you'd rather not manage a GPU, hosted access to every model in this directory is coming via OpenSpeech Cloud.

Can GPT-SoVITS clone voices?

Yes — GPT-SoVITS supports voice cloning from reference audio.

What license is GPT-SoVITS released under?

GPT-SoVITS is released under the MIT license. Always check the repository for the exact terms — some models license code and weights separately.

What languages does GPT-SoVITS support?

GPT-SoVITS supports English, Chinese, Japanese, Korean, Cantonese.

Is there a hosted GPT-SoVITS API?

Yes — OpenSpeech Cloud is launching hosted, pay-per-minute access to GPT-SoVITS and every other model in this directory. No CUDA, Docker, or GPU needed.