# RVC-Boss/GPT-SoVITS

1 min voice data can also be used to train a good TTS model! (few shot voice cloning)

Repository: https://github.com/RVC-Boss/GPT-SoVITS
Canonical: https://ross.abutalabs.com/products/gpt-sovits
Language: Python
License: MIT
License Family: permissive
Topics: text-to-speech, tts, vits, voice-clone, voice-cloneai, voice-cloning
Last push: 2026-08-18T09:16:25+00:00

## Health v2 (maintenance only)
Score: 68/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 28, longevity 68
- inputs: {"age_days": 962, "days_push": 15, "days_rel": 453, "gap_med": 48.0, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 61255, forks 6632 (observed 2026-08-28T04:12:19.000805+00:00)

## What it is
GPT-SoVITS is a Python-based few-shot voice cloning and text-to-speech system with an integrated WebUI. It supports zero-shot TTS from a 5-second sample and fine-tuning with just 1 minute of training data, with cross-lingual inference in English, Japanese, Korean, Cantonese, and Chinese.

## Use cases
- clone a voice from a 1-minute audio sample
- generate speech in a custom voice from text
- fine-tune a TTS model with minimal training data
- do zero-shot text-to-speech from a 5-second voice sample
- transcribe and label training audio with built-in ASR tools
- generate speech in a language different from the training data

## When to choose
- you need high-quality voice cloning with very little training data
- you want a self-hosted TTS pipeline with an integrated WebUI for dataset prep and training
- you need cross-lingual TTS across Chinese, English, Japanese, Korean, or Cantonese

## When to avoid
- you need a lightweight production TTS API without GPU training
- you only need off-the-shelf voices without cloning
- you cannot run GPU workloads locally or in Colab

## Facets
- artifact type: application
- maturity: active
- function: tts, speech-recognition, machine-learning, audio-processing, llm-training
- domain: speech-processing, machine-learning, artificial-intelligence
- platform: python, cross-platform
- tags: voice-cloning, few-shot-learning, zero-shot-tts, webui, sovits, gpt, cross-lingual, audio, docker, web-server, gpu

## Member repositories
- RVC-Boss/GPT-SoVITS (main) score 68

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:19.000805+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:17:34.640988+00:00, confidence not recorded.
  - readme: https://github.com/RVC-Boss/GPT-SoVITS (fetched 2026-08-28T04:12:19.000805+00:00, sha a622a85224a1)
- Data as of 2026-08-30T08:39:29.467469+00:00.