Automatic Speech Recognition on People’s Speech
22.3WERKimi-Audio-Instruct
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Kimi-Audio-InstructModel Size=7B2025.08 | 22.3 | 22.3 | |
| MiDashengLMModel Size=7B2025.08 | 27.8 | 27.8 | |
| Qwen2.5-OmniModel Size=7B2025.08 | 28.6 | 28.6 |
Loading the SOTA2 catalog…