Automatic Speech Recognition on WenetSpeech net
4.68WERFireRedASR
Evaluation Results
| Method | Links | |
|---|---|---|
| FireRedASR2025.12 | 4.68 | |
| Qwen3-Omni-Instruct2026.02 | 4.69 | |
| Index-ASR2025.12 | 5.22 | |
| Kimi Audio2026.02 | 5.37 | |
| Kimi-AudioSize=7B2025.09 | 5.39 | |
| Kimi-Audio2025.12 | 5.45 | |
| Step-Audio 2 mini2025.12 | 5.86 | |
| LongCat-Flash-Omni2026.02 | 6.09 | |
| ERNIE 5.02026.02 | 7.27 | |
| Whisper2025.12 | 10.63 | |
| Gemini-3-Pro2026.02 | 10.94 | |
| ERNIE 5.0-Base2026.02 | 11.57 | |
| GPT-4o-Audio2026.02 | 15.3 | |
| VITA-AudioSize=7B2025.09 | 15.88 | |
| Pretrain+TtTSize=3B2025.09 | 19.85 | |
| TtTSize=3B2025.09 | 44.29 | |
| Qwen2.5-3B (AR)Size=3B2025.09 | 73.64 | |
| Qwen2.5-3B (NAR)Size=3B2025.09 | 111.29 | |
| SpeechGPTSize=7B2025.09 | 124.86 | |
| Mini-OmniSize=0.5B2025.09 | 335.8 |