Spatial Reasoning on SPAR-Bench high-level tasks
51.28High Average ScoreSpaceMind++
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| SpaceMind++Model Type=Open-source2026.05 | 51.28 | 39.78 | 33 | 70 | 48.63 | 53.46 | 51.61 | 53.78 | 27.15 | 45.66 | |
| Doubao-1.5-vision-proModel Type=Closed-source2026.05 | 49.49 | 74.71 | 69.64 | 35.75 | 70.33 | 47.65 | 34.95 | 33.14 | 35.76 | 42.86 | |
| SpaceR-7BModel Type=Open-source2026.05 | 45.61 | 62.35 | 61.61 | 52.25 | 51.92 | 46.81 | 37.9 | 36.05 | 24.83 | 34.17 | |
| InternVL2.5-38BModel Type=Open-source2026.05 | 44.13 | 69.12 | 66.67 | 43.75 | 64.29 | 37.67 | 25.27 | 31.98 | 31.79 | 26.61 | |
| GPT-4oModel Type=Closed-source2026.05 | 43.8 | 65 | 64.88 | 44.75 | 50.82 | 43.21 | 29.84 | 32.56 | 27.81 | 35.29 | |
| Qwen2.5-VL-72BModel Type=Open-source2026.05 | 43.8 | 58.82 | 61.9 | 40.75 | 53.57 | 45.98 | 26.88 | 35.17 | 34.11 | 36.97 | |
| GPT-4.1Model Type=Closed-source2026.05 | 42.93 | 71.76 | 67.26 | 46.25 | 54.95 | 41 | 30.38 | 29.65 | 20.2 | 24.93 | |
| LLaVA-v1.6-7BModel Type=Open-source2026.05 | 20.18 | 51.76 | 7.74 | 6.25 | 32.14 | 6.37 | 39.52 | 10.47 | 21.52 | 5.88 |