Benchmark
EmergentTTS-Eval WER
Scores are percentages. Lower is better.
- Models measured
- 9
- Scores recorded
- 14
- Sources
- 1
- IQ weight
- 0.81
Score distribution
- 12.3%
- Mean
- 11.9%
- Median
- 2.2%
- Std dev
- 10.3%
- 10th percentile
- 16.5%
- 90th percentile
- 10.1%
- Min
This benchmark has a weight of 0.81 for speech models, measured over 7 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Speech 02 HDvoice: English_expressive_narrator· EmergentTTS-Eval data | 10.1% |
| 2 | Gemini 2.5 Flash TTSvoice: Zephyr· EmergentTTS-Eval data | 10.3% |
| 3 | Gemini 2.5 Flash TTSvoice: Zephyr, strong-prompt· EmergentTTS-Eval data | 10.4% |
| 4 | GPT-4o mini TTSvoice: Alloy, baseline· EmergentTTS-Eval data | 10.6% |
| 5 | GPT-4o mini TTSvoice: Alloy, strong-prompt· EmergentTTS-Eval data | 10.8% |
| 6 | GPT-4o mini Audio Previewvoice: Alloy· EmergentTTS-Eval data | 10.9% |
| 7 | Gemini 2.5 Pro TTSvoice: Zephyr, strong-prompt· EmergentTTS-Eval data | 11.8% |
| 8 | GPT-4o Audio Previewvoice: Ballad, strong-prompt· EmergentTTS-Eval data | 11.9% |
| 9 | GPT-4o Audio Previewvoice: Alloy, strong-prompt· EmergentTTS-Eval data | 12% |
| 10 | GPT-4o Audio Previewvoice: Alloy· EmergentTTS-Eval data | 12.4% |
| 11 | GPT-4o mini Audio Previewvoice: Alloy, strong-prompt· EmergentTTS-Eval data | 13.1% |
| 12 | Kokoro 82Mvoice: af_heart· EmergentTTS-Eval data | 13.4% |
| 13 | F5 TTSvoice: basic_ref_en· EmergentTTS-Eval data | 16.5% |
| 14 | Orpheus TTSvoice: Tara· EmergentTTS-Eval data | 17.7% |
Data credits: EmergentTTS-Eval (Boson AI), Apache-2.0 (licence)