compare
CosyVoice vs GPT-SoVITS
The same facts for both, read from GitHub every night, and the relation a person reviewed.
| Fact | CosyVoice | GPT-SoVITS |
|---|---|---|
| Language | Python | Python |
| Licence | Apache-2.0 | MIT |
| Stars | 24k | 62k |
| Latest | v2.0 | 20250606v2pro |
| Last push | 2026-05-25 | 2026-10-06 |
| Release cadence | too few releases | about 53 days between releases |
| Active contributors | 0 commit authors on the default branch in the last 90 days | 2 commit authors on the default branch in the last 90 days |
| Flags | none | none |
How they relate
Both replace ElevenLabs. Alternatives to ElevenLabs →
PartialCosyVoiceMultilingual TTS with zero-shot voice cloning and streaming output.
PartialGPT-SoVITSVoice cloning and fine-tuning from a minute of audio, with a web UI.
CosyVoice
This repository publishes tags, not GitHub releases.
GPT-SoVITS
- 20250606v2pro2025-06-06windows 7z package download|windows 7z package (for 50x0 Nvidia GPU) download
- 20250422v42025-04-22windows 7z package download|windows 7z package (for 50x0 Nvidia GPU) download
- 20250228v32025-02-28windows 7z package download
- 20240821v22024-08-21windows 7z package download
CosyVoice
unsigned The latest release, v2.0, carries no signature GitHub could verify.
Loading the security report
GPT-SoVITS
✓ signed The latest release, 20250606v2pro, carries a signature GitHub verified.
Loading the security report