#1010th of 27 in Voice
real-time-voice-cloning
SV2TTS voice cloning toolbox: speech from a few seconds of reference audio
Live demo ↗
(opens in a new tab)Documentation ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 60.2k
- License
- custom license
- Last commit
- Mar 2026
- Language
- Python
Missed the latest GitHub check; these facts may be older.
Overview
Implements the SV2TTS pipeline: a GE2E speaker encoder, a Tacotron synthesizer and a WaveRNN vocoder. It builds a voice representation from a short audio sample and generates speech from arbitrary text, through a GUI toolbox or a command-line demo. The README says the project is dated and points to newer options such as Chatterbox for higher quality.
Who it is for: Researchers and hobbyists experimenting with classic voice cloning
Strengths
- Implements the full SV2TTS pipeline: encoder, synthesizer and vocoder in one repo
- GUI toolbox and CLI entry points (demo_toolbox.py, demo_cli.py)
- Pretrained models download automatically, with a Hugging Face fallback
- Windows and Linux support, with CPU and CUDA extras via uv
Weaknesses
- README says audio quality is behind current SaaS and newer open-source models
- No tagged releases and no Docker setup
- Requires ffmpeg installed separately
- License is not identified (NOASSERTION)
What it needs
- GPU optional
- Needs ffmpeg, uv, Python
- Models: SV2TTS, GE2E encoder, Tacotron synthesizer, WaveRNN vocoder
Also in Voice
See all 27| Rank | Project | Score |
|---|---|---|
| 1 | voicestudioDesktop app for voice cloning, dubbing and dictation with local speech models | 82 out of 100 |
| 2 | whisper.cppC/C++ port of OpenAI Whisper for local speech-to-text transcription | 76 out of 100 |
| 3 | VoiceboxLocal voice studio for cloning, TTS, dictation and agent speech | 70 out of 100 |
| #4 | buzzOffline desktop app for transcribing and translating audio with Whisper | 67 out of 100 |
| #5 | meetilyDesktop app that records, transcribes and summarizes meetings locally | 66 out of 100 |