Spatial Reasoning on OMNI3D BENCH
44AccuracyVALOR
Evaluation Results
| Method | Links | |
|---|---|---|
| VALORModel Category=Open-Source2025.12 | 44 | |
| VALOR-RLModel Category=Open-Source2025.12 | 43.9 | |
| GPT-4oModel Category=Proprietary2025.12 | 35 | |
| Gemini-2.0-FlashModel Category=Proprietary2025.12 | 34.6 | |
| Claude-3.5-HaikuModel Category=Proprietary2025.12 | 32.5 | |
| Llama3.2-11BModel Category=Open-Source2025.12 | 22.7 | |
| Qwen2.5-VL-7BModel Category=Open-Source2025.12 | 20.7 |