Multimodal Audio Understanding on MMAU v05.15.25 (test)
79.87Sound AccuracyAF-Next-Captioner
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| AF-Next-CaptionerMethod Variant=Captioner2026.04 | 79.87 | 75.3 | 72.13 | 75.76 | |
| AF-Next-InstructMethod Variant=Instruct2026.04 | 78.8 | 74.23 | 69.57 | 74.2 | |
| AF-Next-ThinkMethod Variant=Think2026.04 | 78.7 | 74.73 | 71.5 | 75.01 | |
| Audio Flamingo 3Method Variant=Prior SOTA2026.04 | 75.83 | 74.47 | 66.97 | 72.42 |