Speech Synthesis on LibriSpeech clean (En) (test)
4.87WERF5-TTS
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| F5-TTS2025.10 | 4.87 | 0.7 | 6.97 | 5.93 | 80.85 | 15.03 | 5.47 | 7.35 | 2.94 | 22.33 | |
| RLAIF-SPABackbone=MiniCPM-O 2.62025.10 | 5.8 | 0.72 | 7.23 | 6.53 | 80.85 | 16.06 | 38.8 | 8.82 | 5.88 | 30.08 | |
| MegaTTS32025.10 | 6.9 | 0.71 | 7.1 | 6.12 | 82 | 8.29 | 9.45 | 5.88 | 0 | 21.12 | |
| Chat-TTS2025.10 | 7.85 | 0.66 | 5.99 | 5.75 | 76.43 | 9.84 | 37.31 | 1.47 | 0 | 25.01 | |
| Spark-TTSBackbone=Qwen2.52025.10 | 9.25 | 0.68 | 6.18 | 5.73 | 36.33 | 49.22 | 3.48 | 7.35 | 2.94 | 19.86 |