Text-to-Speech on ZTTS1-Eval In-The-Wild (it)
2.08WERGemini 3.1 Flash
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Gemini 3.1 Flashsupports zero-shot voice cloning=false2026.06 | 2.08 | 3.45 | 19.1 | |
| Eleven Labs V3supports zero-shot voice cloning=false2026.06 | 2.36 | 3.21 | 19.6 | |
| Cartesia Sonic 3.52026.06 | 2.4 | 2.6 | 77.2 | |
| ZONOS2 8B Quality ModeQuality Mode=true2026.06 | 2.51 | 2.63 | 63.8 | |
| Fish S2 Pro2026.06 | 2.79 | 2.54 | 70.4 | |
| Qwen 3 TTS 1.7B2026.06 | 2.85 | 2.76 | 71.9 | |
| Inworld TTS 22026.06 | 3.46 | 2.71 | 56 | |
| VoxCPM 22026.06 | 3.5 | 2.26 | 74.5 | |
| ZONOS2 8BQuality Mode=false2026.06 | 5.61 | 2.32 | 69.2 | |
| Ground truth2026.06 | 6.94 | 2.19 | 78.7 |