Audio & voiceMITPython

GPT-SoVITS

Few-shot voice cloning and TTS with a full training WebUI.

Stars: 62.8K
Forks: 6.7K
Open issues: 896
Last commit: 2026-10-08

Live from the GitHub API · updated 2026-10-15

What is GPT-SoVITS?

GPT-SoVITS clones a voice from about a minute of audio (or zero-shot from a few seconds) and supports Chinese, English, Japanese, Korean and Cantonese. Its WebUI bundles dataset tools — vocal separation, slicing, ASR labeling — and training.

Key features

Zero-shot and few-shot voice cloning
Cross-lingual synthesis
Dataset preparation tools in the WebUI
Windows integrated package

System requirements

Platforms: Windows · Linux · macOS
GPU: NVIDIA recommended for training
VRAM: 6 GB+
Python: 3.10

How to install GPT-SoVITS

Summarized from the project’s README. Projects change quickly — check the GitHub page for the latest instructions.

Windows integrated package

1
Download the integrated package linked in the README and extract it.
2
Double-click go-webui.bat to start the WebUI.

Linux / macOS

Create an environment and run the install script described in the README.

git clone https://github.com/RVC-Boss/GPT-SoVITS.git
cd GPT-SoVITS
conda create -n GPTSoVits python=3.10
conda activate GPTSoVits
bash install.sh

2
Start the WebUI.
```
python webui.py
```

Strengths

Excellent quality with little data
Strong for Chinese and Japanese

Limitations

Complex UI with many steps

GPT-SoVITS alternatives compared

Software	Stars	License	VRAM
GPT-SoVITS Few-shot voice cloning and TTS with a full training WebUI.	62.8K	MIT	6 GB+
Fish Speech Multilingual TTS and voice cloning from Fish Audio.	33.1K	Apache-2.0	4 GB+
F5-TTS Zero-shot voice cloning TTS based on flow matching.	15.4K	MIT	—
RVC WebUI Retrieval-based voice conversion — train a voice model and convert singing or speech.	38.7K	MIT	4 GB+
Coqui TTS Deep-learning text-to-speech toolkit, including the XTTS voice cloning model.	46.1K	MPL-2.0	—
Whisper OpenAI’s open-source speech recognition and translation model.	110.5K	MIT	~1 GB (tiny) to ~10 GB (large)

FAQ

Is GPT-SoVITS free?

Yes — GPT-SoVITS is open-source software released under the MIT license. Check the license (and any model weight licenses) before commercial use.

Does GPT-SoVITS run on Windows, Mac and Linux?

GPT-SoVITS supports: Windows, Linux, macOS.

What hardware do I need for GPT-SoVITS?

NVIDIA recommended for training. VRAM: 6 GB+

How do I install GPT-SoVITS?

1. Download the integrated package linked in the README and extract it. 2. Double-click go-webui.bat to start the WebUI.

What are the best GPT-SoVITS alternatives?

Popular alternatives include Fish Speech, F5-TTS, RVC WebUI, Coqui TTS. See the comparison table on this page.

Open-source AI audio, speech & music tools