Automatic Speech Recognition on VoxPopuli 1.0 (test)
6.2Avg WERMoST
Evaluation results
| Method | Links | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MoST2026.01 | 6.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Phi-4 Multimodal2026.01 | 6.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MinMo2026.01 | 6.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Qwen2-Audio2026.01 | 7.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SeamlessM4T-v22026.01 | 7.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Moshi2026.01 | 8.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LLaMA-Omni22026.01 | 9.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| XLS-RModel capacity=1B2021.11 | 10.6 | 8.8 | 11.5 | 15.1 | 10.8 | 8.2 | 7.7 | 7.3 | 9.6 | 12.5 | 8.7 | 19.5 | 11.3 | 10 | 7.1 | |
| XLS-RModel capacity=0.3B2021.11 | 12.8 | 10.2 | 13 | 19.2 | 12.6 | 9.8 | 9.6 | 7.9 | 11.6 | 14.8 | 10.5 | 24.5 | 14.2 | 12.3 | 8.9 | |
| SpiritLM2026.01 | 14.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AudioLM2026.01 | 15 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VP-10KPre-training=VP-10K, Reference=Wang et al. (2021b)2021.11 | 15.3 | 16.2 | 16.2 | 21.5 | 15.4 | 11 | 12.5 | 9.4 | 12 | 19.7 | 11.8 | 26.1 | 17.1 | 14.1 | 11.1 | |
| SpeechGPT2026.01 | 18.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| No pretrainingPre-training=None, Reference=Wang et al. (2021b)2021.11 | 37.5 | 30 | 29.3 | 45.2 | 30.5 | 31.4 | 25.6 | 27.7 | 27.9 | 38.3 | 27.7 | 96.5 | 41.6 | 40.2 | 32.7 |