Text-to-Speech on EmergentTTS (eval)
6.35Overall WERGemini-2.5-Flash-Preview-TTS
Evaluation Results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gemini-2.5-Flash-Preview-TTSStrong Prompting=true2026.03 | 6.35 | 79.1 | 0.71 | 97.32 | 11.8 | 65.56 | 18.38 | 91.25 | 0.4 | 74.82 | 0.5 | 67.14 | — | |
| Gemini-2.5-Flash-Preview-TTSStrong Prompting=false2026.03 | 6.59 | 78 | 0.6 | 95 | 12.99 | 63.03 | 18.25 | 89.1 | 0.36 | 74.82 | 0.73 | 68.03 | — | |
| minimax/speech-02-hdStrong Prompting=false2026.03 | 6.79 | 40.96 | 0.57 | 41.6 | 14.58 | 31.96 | 17.69 | 33.57 | 0.27 | 52.14 | 0.84 | 45.53 | — | |
| gpt-4o-mini-ttsStrong Prompting=true2026.03 | 7.11 | 61.98 | 0.71 | 67.32 | 12.07 | 63.03 | 21.33 | 63.39 | 0.66 | 54.28 | 0.84 | 61.96 | — | |
| gpt-4o-mini-audio-previewStrong Prompting=false2026.03 | 7.18 | 55.15 | 0.95 | 62.9 | 14.48 | 64.82 | 19.04 | 52.14 | 0.55 | 45.71 | 0.88 | 50.17 | — | |
| Baseline gpt-4o-mini-ttsStrong Prompting=false2026.03 | 7.23 | 50 | 0.72 | — | 13.45 | — | 20.55 | — | 0.42 | — | 1.04 | — | — | |
| Fish Audio S1Strong Prompting=false2026.03 | 7.6 | 36.88 | 0.75 | 24.82 | 15.77 | 53.75 | 20.02 | 37.68 | 0.72 | 43.55 | 0.74 | 24.64 | — | |
| gpt-4o-audio-previewStrong Prompting=true, variant=row 82026.03 | 7.64 | 64.9 | 0.93 | 67.14 | 13.75 | 71.6 | 20.56 | 77.32 | 1.72 | 49.64 | 1.26 | 58.92 | — | |
| 11Labs eleven multilingual v2Strong Prompting=false2026.03 | 7.64 | 36.96 | 0.63 | 35.35 | 14.44 | 36.6 | 21.51 | 52.14 | 0.49 | 28.21 | 1.15 | 32.5 | — | |
| gpt-4o-audio-previewStrong Prompting=true, variant=row 52026.03 | 7.75 | 77.36 | 1.82 | 96.96 | 13.3 | 70 | 21.15 | 89.46 | 1.38 | 62.5 | 1.16 | 68.39 | — | |
| Gemini-2.5-Pro-Preview-TTSStrong Prompting=true2026.03 | 7.91 | 73 | 0.87 | 92.14 | 16.22 | 63.44 | 20.87 | 84.28 | 0.72 | 65.71 | 0.87 | 60 | — | |
| Fish Audio S2Strong Prompting=true2026.03 | 8.15 | 81.88 | 0.52 | 86.61 | 14.74 | 63.39 | 24.58 | 91.61 | 0.38 | 84.41 | 0.51 | 83.39 | — | |
| ResembleAI ChatterboxStrong Prompting=false2026.03 | 8.2 | 27.39 | 1.18 | 28.03 | 17.59 | 24.64 | 20.64 | 17.32 | 0.65 | 51.07 | 0.96 | 15.89 | — | |
| gpt-4o-audio-previewStrong Prompting=false2026.03 | 8.27 | 60.53 | 1.03 | 56.78 | 14.72 | 68.75 | 23.16 | 71.6 | 1.19 | 48.75 | 1.25 | 56.78 | — | |
| HumeAIStrong Prompting=true2026.03 | 8.62 | 49.12 | 0.83 | 71.78 | 21.05 | 41.25 | 19.84 | 42.14 | 0.38 | 42.14 | 0.93 | 48.57 | — | |
| gpt-4o-mini-audio-previewStrong Prompting=true, variant=row 72026.03 | 8.86 | 66.27 | 9.34 | 74.1 | 12.7 | 73.39 | 20.92 | 68.03 | 0.74 | 55.53 | 0.72 | 60.35 | — | |
| KyutAI-TTSStrong Prompting=false2026.03 | 9.1 | 24.28 | 0.83 | 31.78 | 16.47 | 13.39 | 26.41 | 24.28 | 0.61 | 35.17 | 1.19 | 16.78 | — | |
| MiniMax/speech-02-hd⋆Voice=EN-narr2026.06 | 10.02 | 36.6 | — | 40.9 | — | 34.3 | — | 34.3 | — | 47.3 | — | 43.9 | 16.3 | |
| DeepGram Aura-2Strong Prompting=false2026.03 | 10.24 | 25.21 | 3.45 | 17.5 | 21.41 | 15.89 | 23.73 | 20.89 | 1.24 | 43.03 | 1.36 | 28.75 | — | |
| Gemini-2.5-Flash-TTS⋆Voice=Zephyr2026.06 | 10.39 | 70.7 | — | 95.9 | — | 58.5 | — | 91.3 | — | 63 | — | 57.9 | 55.7 | |
| dots.tts (SOAR)Voice=basic_ref_en2026.06 | 10.45 | 47.6 | — | 63.9 | — | 39.4 | — | 52.7 | — | 47 | — | 65.7 | 16.4 | |
| Kokoro-82MStrong Prompting=false2026.03 | 10.49 | 25.46 | 0.71 | 18.03 | 22.17 | 13.21 | 28.37 | 5.89 | 0.56 | 43.39 | 0.65 | 46.78 | — | |
| BASELINE: gpt-4o-mini-ttsVoice=Alloy2026.06 | 10.61 | 50 | — | — | — | — | — | — | — | — | — | — | — | |
| gpt-4o-mini-tts⋆Voice=Alloy2026.06 | 10.76 | 56.3 | — | 59.2 | — | 57.3 | — | 58.8 | — | 52.7 | — | 57.1 | 52.4 | |
| dots.tts (Pretrain)Voice=basic_ref_en2026.06 | 10.86 | 49.2 | — | 72.7 | — | 39.5 | — | 54.7 | — | 48.4 | — | 58.4 | 18 | |
| 11Labs Multilingual v2⋆Voice=Brian2026.06 | 11.19 | 33.9 | — | 30.4 | — | 35.5 | — | 45.5 | — | 39.5 | — | 35.5 | 14.5 | |
| dots.tts (MF4)Voice=basic_ref_en2026.06 | 11.75 | 47.9 | — | 59.8 | — | 36.3 | — | 55.2 | — | 50.5 | — | 64.8 | 16.7 | |
| Gemini-2.5-Pro-TTS⋆Voice=Zephyr2026.06 | 11.79 | 69.3 | — | 86.9 | — | 58.2 | — | 82.3 | — | 61.3 | — | 61.8 | 64.8 | |
| Higgs Audio V2Strong Prompting=false2026.03 | 11.82 | 44.64 | 0.97 | 69.28 | 22.26 | 21.42 | 31.34 | 45.53 | 3.23 | 45.35 | 1.29 | 41.6 | — | |
| VoxCPM2Voice=basic_ref_en2026.06 | 11.84 | 41.1 | — | 42.3 | — | 33.3 | — | 44.1 | — | 53.4 | — | 52.3 | 18.6 | |
| gpt-4o-audio-preview⋆Voice=Ballad2026.06 | 11.87 | 65.2 | — | 88.8 | — | 60.2 | — | 82.1 | — | 57 | — | 59.5 | 40.4 | |
| F5TTSStrong Prompting=false2026.03 | 11.93 | 19.43 | 0.7 | 29.64 | 23.51 | 3.21 | 31.66 | 19.82 | 1.62 | 19.46 | 2.14 | 25 | — | |
| HumeAI⋆2026.06 | 12.85 | 42.7 | — | 61.6 | — | 34.6 | — | 36.9 | — | 43.2 | — | 44.6 | 34.3 | |
| Orpheus TTSStrong Prompting=false2026.03 | 13.63 | 32.56 | 1.81 | 39.06 | 22.31 | 14.64 | 40.94 | 48.57 | 1.48 | 31.07 | 1.63 | 29.46 | — | |
| F5-TTSVoice=basic_ref_en2026.06 | 16.47 | 15.3 | — | 26.8 | — | 1.8 | — | 21.6 | — | 14.8 | — | 23.8 | 1.4 | |
| Qwen3-TTSVoice=basic_ref_en2026.06 | 17.32 | 42.8 | — | 39.8 | — | 25.4 | — | 50.7 | — | 48.9 | — | 60.4 | 30 | |
| Qwen 2.5 OmniStrong Prompting=true2026.03 | 18.34 | 32.94 | 2.41 | 46.6 | 26.77 | 14.46 | 58.44 | 21.25 | 0.87 | 48.92 | 3.47 | 33.75 | — | |
| Qwen3-TTSVoice=Ryan2026.06 | 19.65 | 42.3 | — | 60.5 | — | 17.1 | — | 62.7 | — | 56.4 | — | 43 | 9.8 | |
| Qwen 2.5 OmniStrong Prompting=false2026.03 | 20.02 | 30.25 | 1.22 | 41.07 | 26.98 | 12.5 | 57.48 | 20.89 | 12.77 | 49.1 | 1.66 | 27.67 | — | |
| MiniCPM-oStrong Prompting=false2026.03 | 22.52 | 19.75 | 12.36 | 22.85 | 33.46 | 5.89 | 58.48 | 21.6 | 5.21 | 22.14 | 3.08 | 26.25 | — |