Text-to-Speech on X-Voice (test)
1.37WERFish Audio S2
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Fish Audio S2Language=Korean2026.05 | 1.37 | 70.2 | |
| OmniVoiceLanguage=Portuguese2026.05 | 2.11 | 72 | |
| Qwen3-TTSLanguage=Spanish2026.05 | 2.11 | 74.9 | |
| OmniVoiceLanguage=Indonesian2026.05 | 2.15 | 68.2 | |
| Qwen3-TTSLanguage=Chinese2026.05 | 2.16 | 72.8 | |
| OmniVoiceLanguage=Dutch2026.05 | 2.18 | 71.3 | |
| OmniVoiceLanguage=Chinese2026.05 | 2.23 | 73.6 | |
| OmniVoiceLanguage=Spanish2026.05 | 2.28 | 75.9 | |
| X-Voices2Language=English, Fine-tuning without reference transcripts=true2026.05 | 2.29 | 54.7 | |
| X-Voices1Language=English, Textual conditioning=true2026.05 | 2.36 | 58.6 | |
| X-Voices2Language=Korean, Fine-tuning without reference transcripts=true2026.05 | 2.4 | 73.1 | |
| Ground TruthLanguage=Chinese2026.05 | 2.41 | 72.3 | |
| X-Voices1Language=Korean, Textual conditioning=true2026.05 | 2.42 | 72.3 | |
| OmniVoiceLanguage=English2026.05 | 2.44 | 71.9 | |
| X-Voices2Language=Indonesian, Fine-tuning without reference transcripts=true2026.05 | 2.53 | 65.1 | |
| Fish Audio S2Language=Chinese2026.05 | 2.57 | 68.6 | |
| Qwen3-TTSLanguage=German2026.05 | 2.64 | 74.2 | |
| X-Voices1Language=Russian, Textual conditioning=true2026.05 | 2.68 | 71.4 | |
| Fish Audio S2Language=Portuguese2026.05 | 2.71 | 65.8 | |
| X-Voices2Language=Russian, Fine-tuning without reference transcripts=true2026.05 | 2.74 | 72.3 | |
| Qwen3-TTSLanguage=Portuguese2026.05 | 2.78 | 71.1 | |
| OmniVoiceLanguage=German2026.05 | 2.8 | 76.3 | |
| X-Voices1Language=Spanish, Textual conditioning=true2026.05 | 2.83 | 69.5 | |
| X-Voices1Language=Chinese, Textual conditioning=true2026.05 | 2.86 | 69.8 | |
| X-Voices2Language=Chinese, Fine-tuning without reference transcripts=true2026.05 | 2.87 | 70 | |
| X-Voices2Language=Spanish, Fine-tuning without reference transcripts=true2026.05 | 2.89 | 69.3 | |
| OmniVoiceLanguage=Thai2026.05 | 2.9 | 69.8 | |
| Fish Audio S2Language=Spanish2026.05 | 2.9 | 69.4 | |
| MOSS-TTSLanguage=Chinese2026.05 | 2.91 | 72.2 | |
| Qwen3-TTSLanguage=Italian2026.05 | 2.93 | 74.6 | |
| X-Voices1Language=Indonesian, Textual conditioning=true2026.05 | 2.98 | 64.4 | |
| Fish Audio S2Language=English2026.05 | 3.01 | 62.2 | |
| OmniVoiceLanguage=Polish2026.05 | 3.06 | 71.3 | |
| OmniVoiceLanguage=Russian2026.05 | 3.12 | 74.4 | |
| OmniVoiceLanguage=Italian2026.05 | 3.13 | 75.7 | |
| X-Voices2Language=Dutch, Fine-tuning without reference transcripts=true2026.05 | 3.13 | 66.9 | |
| Qwen3-TTSLanguage=Russian2026.05 | 3.16 | 74.2 | |
| X-Voices1Language=Dutch, Textual conditioning=true2026.05 | 3.19 | 66.7 | |
| X-Voices2Language=Portuguese, Fine-tuning without reference transcripts=true2026.05 | 3.29 | 66.1 | |
| X-Voices1Language=Polish, Textual conditioning=true2026.05 | 3.3 | 67.9 | |
| Ground TruthLanguage=Spanish2026.05 | 3.32 | 76.2 | |
| Ground TruthLanguage=Vietnamese2026.05 | 3.37 | 74.1 | |
| Ground TruthLanguage=Portuguese2026.05 | 3.39 | 72.4 | |
| X-Voices1Language=Portuguese, Textual conditioning=true2026.05 | 3.41 | 66.5 | |
| OmniVoiceLanguage=Korean2026.05 | 3.42 | 74.8 | |
| Fish Audio S2Language=German2026.05 | 3.45 | 68.9 | |
| Ground TruthLanguage=German2026.05 | 3.51 | 76.5 | |
| Fish Audio S2Language=Russian2026.05 | 3.52 | 68.9 | |
| MOSS-TTSLanguage=Korean2026.05 | 3.54 | 72.8 | |
| MOSS-TTSLanguage=English2026.05 | 3.54 | 66.2 | |
| LEMAS-TTSLanguage=Portuguese2026.05 | 3.55 | 64.7 | |
| Qwen3-TTSLanguage=Korean2026.05 | 3.64 | 73.8 | |
| Ground TruthLanguage=Indonesian2026.05 | 3.67 | 67 | |
| X-Voices2Language=Vietnamese, Fine-tuning without reference transcripts=true2026.05 | 3.67 | 68.6 | |
| X-Voices2Language=Polish, Fine-tuning without reference transcripts=true2026.05 | 3.71 | 68.3 | |
| Ground TruthLanguage=Russian2026.05 | 3.75 | 74.3 | |
| X-Voices1Language=German, Textual conditioning=true2026.05 | 3.76 | 69.8 | |
| LEMAS-TTSLanguage=Russian2026.05 | 3.8 | 67.6 | |
| MOSS-TTSLanguage=German2026.05 | 3.85 | 71.8 | |
| MOSS-TTSLanguage=Spanish2026.05 | 3.88 | 72.6 | |
| Qwen3-TTSLanguage=English2026.05 | 3.89 | 69.7 | |
| X-Voices2Language=Italian, Fine-tuning without reference transcripts=true2026.05 | 3.89 | 70.5 | |
| X-Voices2Language=German, Fine-tuning without reference transcripts=true2026.05 | 3.91 | 69.8 | |
| X-Voices1Language=Italian, Textual conditioning=true2026.05 | 3.95 | 70.3 | |
| Ground TruthLanguage=Thai2026.05 | 3.98 | 69.4 | |
| Ground TruthLanguage=Dutch2026.05 | 4.1 | 72.5 | |
| LEMAS-TTSLanguage=English2026.05 | 4.15 | 56 | |
| X-Voices1Language=Vietnamese, Textual conditioning=true2026.05 | 4.25 | 67.7 | |
| OmniVoiceLanguage=Vietnamese2026.05 | 4.38 | 72.7 | |
| Fish Audio S2Language=Dutch2026.05 | 4.38 | 65 | |
| LEMAS-TTSLanguage=Indonesian2026.05 | 4.4 | 60.4 | |
| X-Voices1Language=Finnish, Textual conditioning=true2026.05 | 4.41 | 71.9 | |
| Ground TruthLanguage=Italian2026.05 | 4.44 | 76 | |
| X-Voices2Language=Finnish, Fine-tuning without reference transcripts=true2026.05 | 4.47 | 72.2 | |
| Ground TruthLanguage=Korean2026.05 | 4.54 | 75.9 | |
| OmniVoiceLanguage=Croatian2026.05 | 4.56 | 80.1 | |
| OmniVoiceLanguage=Czech2026.05 | 4.58 | 73.6 | |
| Fish Audio S2Language=Italian2026.05 | 4.6 | 69.2 | |
| LEMAS-TTSLanguage=Spanish2026.05 | 4.6 | 66.7 | |
| Ground TruthLanguage=English2026.05 | 4.63 | 73 | |
| X-Voices1Language=Croatian, Textual conditioning=true2026.05 | 4.81 | 78.2 | |
| MOSS-TTSLanguage=Russian2026.05 | 4.82 | 71.8 | |
| X-Voices2Language=Croatian, Fine-tuning without reference transcripts=true2026.05 | 4.84 | 79 | |
| X-Voices2Language=Czech, Fine-tuning without reference transcripts=true2026.05 | 4.84 | 70.6 | |
| X-Voices1Language=Czech, Textual conditioning=true2026.05 | 4.96 | 70.2 | |
| LEMAS-TTSLanguage=Vietnamese2026.05 | 4.97 | 64.2 | |
| OmniVoiceLanguage=Swedish2026.05 | 5.06 | 73.4 | |
| Ground TruthLanguage=Polish2026.05 | 5.12 | 72.6 | |
| OmniVoiceLanguage=Finnish2026.05 | 5.31 | 75.4 | |
| LEMAS-TTSLanguage=Italian2026.05 | 5.39 | 65.5 | |
| X-Voices2Language=Hungarian, Fine-tuning without reference transcripts=true2026.05 | 5.42 | 70.1 | |
| X-Voices1Language=Thai, Textual conditioning=true2026.05 | 5.47 | 65.4 | |
| X-Voices1Language=Hungarian, Textual conditioning=true2026.05 | 5.56 | 70 | |
| X-Voices2Language=Thai, Fine-tuning without reference transcripts=true2026.05 | 5.64 | 67.1 | |
| OmniVoiceLanguage=Japanese2026.05 | 5.98 | 72.4 | |
| Fish Audio S2Language=Japanese2026.05 | 6.06 | 65.8 | |
| LEMAS-TTSLanguage=Chinese2026.05 | 6.07 | 65.5 | |
| Fish Audio S2Language=Indonesian2026.05 | 6.11 | 59.6 | |
| Fish Audio S2Language=Polish2026.05 | 6.12 | 65.7 | |
| MOSS-TTSLanguage=Portuguese2026.05 | 6.19 | 67.9 |