Visual Reasoning on V* (test)
92.2Overall ScoreDRIM
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DRIM2025.12 | 92.2 | 91.3 | 93.4 | |
| Mini-o3Source=Reproduced through available model weights2025.12 | 91.1 | 90.4 | 92.1 | |
| DeepEyesSource=Reproduced through available model weights2025.12 | 89.5 | 90.4 | 88.2 | |
| ThymeSource=Reported in original or other relevant paper2025.12 | 82.2 | 83.5 | 80.3 | |
| DeepEyesv2Source=Reported in original or other relevant paper2025.12 | 81.8 | — | — | |
| Gemini 2.5 Pro2025.12 | 81.7 | 87.8 | 72.4 | |
| DyFoSource=Reported in original or other relevant paper2025.12 | 81.2 | 80 | 82.9 | |
| Ovis2.5-9B2025.12 | 80.6 | 81.7 | 78.9 | |
| SEALSource=Reported in original or other relevant paper2025.12 | 75.4 | 74.8 | 76.3 | |
| GPT-52025.12 | 74.3 | 72.2 | 77.6 | |
| LLaVA-OneVisionSource=Reported in original or other relevant paper2025.12 | 70.9 | — | — |