Spoken Dialogue on Big Bench Audio (test)
72.9S2T AccuracyMiMo-Audio-7B-Instruct
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MiMo-Audio-7B-Instruct2025.12 | 72.9 | 60.2 | |
| gpt-4o-audio-preview-2024-12-172025.12 | 70.2 | 67.2 | |
| Kimi-Audio-Instruct2025.12 | 59.4 | 51 | |
| Qwen2.5-Omni2025.12 | 54.2 | 53.6 | |
| Step-Audio2-mini2025.12 | 50.9 | 47.5 | |
| GLM-4-Voice2025.12 | 44.8 | 42.7 |