# voicepaw/so-vits-svc-fork

so-vits-svc fork with realtime support, improved interface and more features.

Repository: https://github.com/voicepaw/so-vits-svc-fork
Canonical: https://ross.abutalabs.com/products/so-vits-svc-fork
Language: Python
License: NOASSERTION
License Family: other
Topics: sovits, vits, voice-conversion, so-vits-svc, hubert, softvc, realtime, voice-changer, deep-learning, pytorch, speech-synthesis, contentvec, gan, lightning, pytorch-lightning
Last push: 2026-08-26T20:10:07+00:00

## Health v2 (maintenance only)
Score: 82/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 56, longevity 90
- inputs: {"age_days": 1267, "days_push": 7, "days_rel": 212, "gap_med": 45, "n_releases_24m": 4}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 9327, forks 1225 (observed 2026-08-28T04:10:30.546352+00:00)

## What it is
A fork of so-vits-svc providing singing voice conversion with realtime support and an improved interface, built on PyTorch and PyTorch Lightning. It trains and runs SoftVC VITS voice conversion models compatible with so-vits-svc 4.0/4.1 checkpoints.

## Use cases
- convert my singing voice to another singer's voice
- realtime voice changer for streaming
- train a voice conversion model from audio samples
- make AI cover songs with a custom voice model
- run offline singing voice conversion on gpu

## When to choose
- you need realtime singing voice conversion with so-vits-svc 4.0-compatible models
- you already have so-vits-svc v1 models and want a better interface

## When to avoid
- you want an actively maintained project - it is explicitly no longer maintained
- you have 4.1 or other incompatible model formats
- you want the latest voice conversion technology - alternatives like RVC/Applio are recommended

## Facets
- artifact type: application
- maturity: abandoned
- function: speech-recognition, audio-processing, machine-learning, deep-learning, llm-training
- domain: speech-processing, machine-learning, deep-learning
- platform: python, cross-platform, cli
- tags: voice-conversion, singing-voice-conversion, vits, so-vits-svc, realtime-voice-changer, hubert, contentvec, pytorch-lightning, speech-synthesis, voice-changer, audio, gpu

## Member repositories
- voicepaw/so-vits-svc-fork (main) score 82

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:30.546352+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 2026-08-29T17:22:29.335404+00:00, confidence not recorded.
  - readme: https://github.com/voicepaw/so-vits-svc-fork (fetched 2026-08-28T04:10:30.546352+00:00, sha f8ef93ef6ffc)
  - registry_pypi: https://pypi.org/pypi/so-vits-svc-fork/json (fetched 2026-08-29T08:22:25.869231+00:00, sha fb7845b9929e)
- Data as of 2026-08-30T08:39:29.467469+00:00.