Spatial Reasoning on VLM4D-Real Overall
82Accuracy (%)Reasmory
Evaluation Results
| Method | Links | |
|---|---|---|
| ReasmoryBase Model=Gemini-3-flash, Category=Test-time Scaling Methods2026.05 | 82 | |
| Gemini-3-flash2026.05 | 76 | |
| ReasmoryBase Model=GPT-5-mini, Category=Test-time Scaling Methods2026.05 | 72.7 | |
| Think3DBase Model=Gemini-3-flash, Category=Test-time Scaling Methods2026.05 | 69 | |
| Think3DBase Model=GPT-5-mini, Category=Test-time Scaling Methods2026.05 | 67.6 | |
| GPT-5-mini2026.05 | 65.3 | |
| MindjourneyBase Model=Gemini-3-flash, Category=Test-time Scaling Methods2026.05 | 63 | |
| MindjourneyBase Model=GPT-5-mini, Category=Test-time Scaling Methods2026.05 | 50 | |
| SpatialLadder-3BCategory=Spatial Models2026.05 | 39 | |
| SpaceOM-3BCategory=Spatial Models2026.05 | 38.5 | |
| Spatial-MLLM-4BCategory=Spatial Models2026.05 | 30 |