Spoken Question Answering on MuSiQue v1.0 (test)
53.99Exact Match (EM)AEG (with LFE)
Evaluation Results
| Method | Links | |
|---|---|---|
| AEG (with LFE)Model=GPT-4o Audio2026.03 | 53.99 | |
| AEG (with LFE)Model=LongCat-Flash-Omni2026.03 | 53.99 | |
| AEG (w/o LFE)Model=GPT-4o Audio2026.03 | 53.7 | |
| AEG (w/o LFE)Model=LongCat-Flash-Omni2026.03 | 52.38 | |
| baselineModel=GPT-4o Audio2026.03 | 51.51 | |
| baselineModel=LongCat-Flash-Omni2026.03 | 49.57 | |
| AEG (with LFE)Model=Qwen3-Omni-30B-A3B2026.03 | 48.61 | |
| AEG (w/o LFE)Model=Qwen3-Omni-30B-A3B2026.03 | 47.62 | |
| baselineModel=Qwen3-Omni-30B-A3B2026.03 | 45.88 | |
| AEG (with LFE)Model=Qwen3-Omni Flash2026.03 | 45.72 | |
| AEG (w/o LFE)Model=Qwen3-Omni Flash2026.03 | 44.93 | |
| baselineModel=Qwen3-Omni Flash2026.03 | 44.72 |