Text-to-Speech on ZTTS1-Eval Clean 1.0 (test)
1.94WER (en)Qwen 3 TTS 1.7B
Evaluation results
| Method | Links | |||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Qwen 3 TTS 1.7Bmode=zero-shot, parameters=1.7B2026.06 | 1.94 | 3.86 | 68.3 | 1.26 | 3.33 | 60.2 | 2.91 | 3.71 | 79.7 | 6.04 | 3.16 | 75.8 | 2.78 | 3.72 | 69.7 | 1.98 | 3.43 | 69.2 | 3.08 | 3.64 | 72 | 2.19 | 3.38 | 75.2 | 3.78 | 3.71 | 81 | 2.85 | 3.53 | 79.3 | 4.25 | 3.49 | 74.1 | |
| Eleven Labs V3mode=zero-shot, supports_cloning=false2026.06 | 2.35 | 3.59 | 7.2 | 0.75 | 3.53 | 6.8 | 4.2 | 3.31 | 36.9 | 8.31 | 3.37 | 29.4 | 3.79 | 3.54 | 17.1 | 3.24 | 3.25 | 18 | 3.62 | 3.29 | 21.9 | 2.7 | 3.19 | 21.9 | 4.26 | 3.45 | 36.3 | 4.57 | 3.36 | 38.6 | 5.63 | 3.15 | 25.4 | |
| Gemini 3.1 Flashmode=zero-shot, supports_cloning=false2026.06 | 2.5 | 3.87 | 12.7 | 0.8 | 3.79 | 9.8 | 4.09 | 3.54 | 36.6 | 5.63 | 3.37 | 28.6 | 3.61 | 3.71 | 17.7 | 4.41 | 3.43 | 19.5 | 4.77 | 3.52 | 20.9 | 3.24 | 3.39 | 20.4 | 3.79 | 3.65 | 33.6 | 3.06 | 3.51 | 36.2 | 6.58 | 3.37 | 23.5 | |
| Cartesia Sonic 3.5mode=zero-shot2026.06 | 2.56 | 3.62 | 79.9 | 0.86 | 3.19 | 67.1 | 4.53 | 3.17 | 85.4 | 7.66 | 3.01 | 80 | 3.14 | 3.18 | 82.8 | 3.21 | 2.97 | 85.9 | 3.41 | 3.29 | 83.1 | 3.05 | 2.9 | 83.4 | 3.65 | 3.36 | 87.5 | 3.24 | 3.11 | 86.4 | 6.01 | 3.05 | 84.2 | |
| ZONOS2 8Bmode=zero-shot, parameters=8B2026.06 | 2.76 | 3.4 | 78.6 | 15.04 | 2.57 | 62.1 | 15.62 | 3.1 | 73.3 | 26.23 | 2.29 | 74.6 | 5.67 | 3.23 | 78.3 | 4.78 | 2.96 | 79.4 | 13.18 | 3.27 | 70 | 6.53 | 2.89 | 77.7 | 8.27 | 3.23 | 81.3 | 18.96 | 3.13 | 73.3 | 8.26 | 2.97 | 80.2 | |
| Inworld TTS 2mode=zero-shot2026.06 | 3.14 | 3.53 | 65.8 | 1.96 | 3.15 | 54.8 | 4.04 | 3.14 | 74.8 | 7.86 | 2.94 | 67.1 | 4.05 | 3.27 | 68.4 | 4.38 | 3.05 | 73 | 4.95 | 3.33 | 68.5 | 3.98 | 2.95 | 70.4 | 5.63 | 3.27 | 77.7 | 3.13 | 3.11 | 78.3 | 7.08 | 3.01 | 69.8 | |
| Fish S2 Promode=zero-shot2026.06 | 3.6 | 3.47 | 76.9 | 5 | 3.06 | 64.7 | 4.33 | 3.23 | 82.9 | 7.95 | 3 | 76 | 3.04 | 3.26 | 78.5 | 3.73 | 3.01 | 82.8 | 5.9 | 3.22 | 78 | 4.17 | 3.02 | 79.3 | 4.2 | 3.37 | 85.4 | 3.92 | 3.18 | 83.9 | 7.74 | 3.09 | 80.6 | |
| ZONOS2 8B Quality Modemode=zero-shot, parameters=8B, quality_mode=true2026.06 | 3.99 | 3.47 | 74.4 | 2.68 | 2.94 | 58.2 | 6.73 | 3.21 | 81.1 | 14.41 | 2.65 | 73.4 | 3.74 | 3.36 | 76.2 | 3.25 | 2.94 | 79 | 4.3 | 3.31 | 75.9 | 3.52 | 2.92 | 78 | 7.67 | 3.31 | 82 | 4.14 | 3.18 | 83 | 6.99 | 3.02 | 79.4 | |
| VoxCPM 2mode=zero-shot2026.06 | 4.23 | 3.51 | 65.2 | 0.94 | 2.82 | 66.8 | 5.01 | 3.12 | 77.3 | 5.21 | 2.55 | 77.5 | 4.84 | 3.07 | 79.7 | 4.44 | 2.85 | 75.4 | 5.1 | 3.25 | 75.8 | 5.8 | 2.71 | 80.8 | 4.92 | 3.23 | 83.3 | 3.99 | 3.04 | 83.1 | 7.43 | 2.83 | 80.4 |