Audio Reasoning on MMAR Agent Track
77.4AccuracyTeam B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Team BModel Category=Agent-based2026.03 | 77.4 | 66.23 | |
| TalTechModel Category=Agent-based2026.03 | 76.9 | 69.83 | |
| Team CModel Category=Agent-based2026.03 | 75.1 | 66.09 | |
| Gemini 2.5 ProModel Category=Single-model2026.03 | 74.7 | — | |
| Team DModel Category=Agent-based2026.03 | 72.2 | 64.61 | |
| Team EModel Category=Agent-based2026.03 | 71 | 63 | |
| Qwen3-Omni-ThinkingModel Category=Single-model2026.03 | 66.4 | — | |
| GPT-4o AudioModel Category=Single-model2026.03 | 63.5 | — |