Spatial Understanding on MindCube Tiny
87Rotation Scoredoubao-seed-1.6
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| doubao-seed-1.6Model Category=Closed-source Models2025.10 | 87 | 35.8 | 38 | 46.1 | |
| o3-2025-04-16Model Category=Closed-source Models2025.10 | 86.5 | 42.7 | 66 | 56.6 | |
| gemini-2.5-proModel Category=Closed-source Models2025.10 | 84 | 39.7 | 56.8 | 52.2 | |
| 3DThinkerBase VLM=Qwen2.5-VL-72B, Training Stage=S1+S22025.10 | 57 | 83.7 | 77.6 | 77.1 | |
| 3DThinkerBase VLM=InternVL3-78B, Training Stage=S1+S22025.10 | 57 | 86.2 | 78.8 | 78.9 | |
| 3DThinkerBase VLM=Qwen2.5-VL-32B, Training Stage=S1+S22025.10 | 56.5 | 83.2 | 77.2 | 76.7 | |
| 3DThinkerBase VLM=Qwen2.5-VL-3B, Training Stage=S1+S22025.10 | 55.5 | 81.8 | 75.2 | 75.2 | |
| 3DThinkerBase VLM=Qwen2.5-VL-7B, Training Stage=S1+S22025.10 | 55 | 83 | 76 | 76 | |
| 3DThinkerBase VLM=InternVL3-8B, Training Stage=S1+S22025.10 | 55 | 82.5 | 79.2 | 76.5 | |
| 3DThinkerBase VLM=InternVL3-14B, Training Stage=S1+S22025.10 | 54.5 | 84.3 | 77.6 | 77 | |
| 3DThinkerBase VLM=InternVL3-38B, Training Stage=S1+S22025.10 | 53.5 | 85.2 | 78 | 77.4 | |
| claude-sonnet-4Model Category=Closed-source Models2025.10 | 49.5 | 42.2 | 12.8 | 36.6 | |
| 3DThinker-S1+S2Model Scale=8B, Training Stage=S1 + S2, Base Model=LLaVA-OneVision-1.52025.10 | 49 | 68.2 | 64.8 | 63.7 | |
| 3DThinker-S1+S2Model Scale=4B, Training Stage=S1 + S2, Base Model=LLaVA-OneVision-1.52025.10 | 48 | 67.5 | 65.2 | 63.2 | |
| gpt-4.1Model Category=Closed-source Models2025.10 | 45.5 | 44.2 | 47.2 | 45.1 | |
| 3DThinkerBase VLM=Qwen2.5-VL-32B, Training Stage=S12025.10 | 45 | 66.8 | 77.2 | 65.1 | |
| 3DThinkerBase VLM=Qwen2.5-VL-3B, Training Stage=S12025.10 | 44 | 64.8 | 72.4 | 62.7 | |
| 3DThinkerBase VLM=Qwen2.5-VL-7B, Training Stage=S12025.10 | 43.5 | 66.3 | 76.4 | 64.4 | |
| 3DThinkerBase VLM=InternVL3-78B, Training Stage=S12025.10 | 43.5 | 69 | 77.2 | 66.1 | |
| 3DThinkerBase VLM=InternVL3-8B, Training Stage=S12025.10 | 43 | 66.8 | 79.2 | 65.2 | |
| 3DThinker-S1Model Scale=8B, Training Stage=S1, Base Model=LLaVA-OneVision-1.52025.10 | 43 | 57.8 | 64.8 | 56.7 | |
| 3DThinkerBase VLM=Qwen2.5-VL-72B, Training Stage=S12025.10 | 42.5 | 68 | 73.6 | 64.5 | |
| 3DThinkerBase VLM=InternVL3-14B, Training Stage=S12025.10 | 42 | 68.3 | 77.2 | 65.4 | |
| 3DThinker-S1Model Scale=4B, Training Stage=S1, Base Model=LLaVA-OneVision-1.52025.10 | 41.5 | 59.8 | 66 | 57.8 | |
| Qwen2.5-VL-72BFamily=Qwen2.5-VL Family2025.10 | 40 | 42.5 | 44.4 | 42.5 | |
| Qwen2.5-VL-32BFamily=Qwen2.5-VL Family2025.10 | 39.5 | 34.5 | 43.6 | 37.6 | |
| 3DThinkerBase VLM=InternVL3-38B, Training Stage=S12025.10 | 39 | 68 | 76.8 | 64.6 | |
| InternVL3-78BFamily=InternVL3 Family2025.10 | 38.5 | 50.5 | 57.4 | 49.9 | |
| Qwen2.5-VL-3BFamily=Qwen2.5-VL Family2025.10 | 37.4 | 33.3 | 30.3 | 33.2 | |
| gpt-4o-2024-11-20Model Category=Closed-source Models2025.10 | 37 | 44.8 | 56.4 | 46.1 | |
| InternVL3-8BFamily=InternVL3 Family2025.10 | 37 | 40.3 | 63.2 | 45.1 | |
| Qwen2.5-VL-7BFamily=Qwen2.5-VL Family2025.10 | 36.5 | 32.5 | 38.4 | 34.7 | |
| InternVL3-14BFamily=InternVL3 Family2025.10 | 36 | 48 | 55.6 | 47.5 | |
| LLaVA-OneVision-1.5-8BModel Scale=8B, Training Stage=Base2025.10 | 34.5 | 34.7 | 48.4 | 37.9 | |
| LLaVA-OneVision-1.5-4BModel Scale=4B, Training Stage=Base2025.10 | 33.5 | 38 | 49.2 | 39.8 | |
| InternVL3-38BFamily=InternVL3 Family2025.10 | 32.5 | 48.5 | 56 | 47.2 | |
| glm-4.5vModel Category=Closed-source Models2025.10 | 28 | 43 | 33.2 | 37.8 |