Text-to-Speech on CosyVoice en 3 (test)
3.931UTMOSQwen3-TTS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen3-TTSSize=1.7B, Emilia=×2026.05 | 3.931 | 3.406 | 0.67 | |
| CosyVoice 2Size=0.5B, Emilia=×2026.05 | 3.85 | 6.2 | 0.614 | |
| CosyVoice 3Size=0.5B, Emilia=×2026.05 | 3.719 | 4.124 | 0.676 | |
| Qwen3-TTSSize=0.6B, Emilia=×2026.05 | 3.68 | 3.636 | 0.641 | |
| OmniVoiceSize=0.6B, Emilia=×2026.05 | 3.619 | 3.182 | 0.705 | |
| Spark-TTSSize=0.5B, Emilia=✓2026.05 | 3.587 | 6.84 | 0.499 | |
| IndexTTS2Size=1.5B, Emilia=INCL.2026.05 | 3.372 | 3.987 | 0.664 | |
| GibbsTTSSize=0.4B, Emilia=✓2026.05 | 3.238 | 4.11 | 0.691 | |
| F5-TTSSize=0.3B, Emilia=✓2026.05 | 3.188 | 6.582 | 0.624 | |
| MaskGCTSize=1.5B, Emilia=✓2026.05 | 3.032 | 7.237 | 0.69 |