Spatial Reasoning on SpatialScore
77.7Original ScoreOurs
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Ours2026.04 | 77.7 | 68.4 | 77.6 | |
| Qwen3-VL-8BModel Type=Open-source Models2026.04 | 73.6 | 57.9 | 73.5 | |
| Gemini-2.5-FlashModel Type=Closed-source Models2026.04 | 73.6 | 57.9 | 73.5 | |
| InternVL3.5-8BModel Type=Open-source Models2026.04 | 72.2 | 47.4 | 72.1 | |
| GPT-5.2Model Type=Closed-source Models, Scale=~100B active parameters, Training Data=≥30T tokens2026.04 | 71.5 | 73.7 | 71.5 | |
| Cambrian-1-8BModel Type=Open-source Models2026.04 | 62.6 | 57.9 | 62.6 | |
| VLM3R-7BModel Type=Spatially Specialized Models, Evaluation Mode=Exclusively processes video inputs (single-frame videos)2026.04 | 62.6 | 63.2 | 62.6 | |
| SpaceMantis-8BModel Type=Spatially Specialized Models2026.04 | 62.3 | 52.6 | 62.3 | |
| SpaceQwen2.5-VL-3BModel Type=Spatially Specialized Models2026.04 | 61.9 | 42.1 | 61.7 | |
| LLaVa-1.6-Mistral-7BModel Type=Open-source Models2026.04 | 60.7 | 68.4 | 60.8 | |
| SpatialLLM-8BModel Type=Spatially Specialized Models2026.04 | 57.8 | 52.6 | 57.8 | |
| Spatial-MLLM-7BModel Type=Spatially Specialized Models, Evaluation Mode=Exclusively processes video inputs (single-frame videos)2026.04 | 50.4 | 36.8 | 50.3 | |
| LLaVA-OneVision-1.5-8BModel Type=Open-source Models2026.04 | 48.5 | 10.5 | 48.2 |