Room Navigation on Room Navigation Visual Appearance - VA
62Success RateSceneDiver
Evaluation Results
| Method | Links | |
|---|---|---|
| SceneDiverBackbone=gemini-2.5-flash, Model Category=Closed-Source2026.06 | 62 | |
| Multi-ResBackbone=gemini-2.5-flash, Model Category=Closed-Source2026.06 | 58 | |
| ThinkingBackbone=gemini-2.5-flash, Model Category=Closed-Source2026.06 | 56.7 | |
| SoMBackbone=gemini-2.5-flash, Model Category=Closed-Source2026.06 | 56 | |
| Base ModelBackbone=gemini-2.5-flash, Model Category=Closed-Source2026.06 | 55.3 | |
| SceneDiverBackbone=Qwen2.5-VL-32B-Instruct-AWQ, Model Category=Open-Source2026.06 | 49.3 | |
| SceneDiverBackbone=gpt-4o-mini, Model Category=Closed-Source2026.06 | 49.3 | |
| VCDBackbone=Qwen2.5-VL-32B-Instruct-AWQ, Model Category=Open-Source2026.06 | 46 | |
| Multi-ResBackbone=Qwen2.5-VL-32B-Instruct-AWQ, Model Category=Open-Source2026.06 | 45.3 | |
| SoMBackbone=Qwen2.5-VL-32B-Instruct-AWQ, Model Category=Open-Source2026.06 | 44.7 | |
| SceneDiverBackbone=doubao-seed-1.6-flash, Model Category=Closed-Source2026.06 | 44.7 | |
| Base ModelBackbone=Qwen2.5-VL-32B-Instruct-AWQ, Model Category=Open-Source2026.06 | 43.3 | |
| Multi-ResBackbone=gpt-4o-mini, Model Category=Closed-Source2026.06 | 43.3 | |
| SoMBackbone=gpt-4o-mini, Model Category=Closed-Source2026.06 | 38.7 | |
| Base ModelBackbone=gpt-4o-mini, Model Category=Closed-Source2026.06 | 37.3 | |
| Multi-ResBackbone=doubao-seed-1.6-flash, Model Category=Closed-Source2026.06 | 37.3 | |
| ThinkingBackbone=doubao-seed-1.6-flash, Model Category=Closed-Source2026.06 | 36.7 | |
| SceneDiverBackbone=Qwen2.5-VL-7B-Instruct-AWQ, Model Category=Open-Source2026.06 | 35.3 | |
| SoMBackbone=doubao-seed-1.6-flash, Model Category=Closed-Source2026.06 | 34 | |
| VCDBackbone=Qwen2.5-VL-7B-Instruct-AWQ, Model Category=Open-Source2026.06 | 33.3 | |
| Base ModelBackbone=doubao-seed-1.6-flash, Model Category=Closed-Source2026.06 | 32.7 | |
| SoMBackbone=Qwen2.5-VL-7B-Instruct-AWQ, Model Category=Open-Source2026.06 | 29.3 | |
| Multi-ResBackbone=Qwen2.5-VL-7B-Instruct-AWQ, Model Category=Open-Source2026.06 | 29.3 | |
| Base ModelBackbone=Qwen2.5-VL-7B-Instruct-AWQ, Model Category=Open-Source2026.06 | 27.3 | |
| SceneDiverBackbone=InternVL2.5-8B, Model Category=Open-Source2026.06 | 25.3 | |
| SoMBackbone=InternVL2.5-8B, Model Category=Open-Source2026.06 | 24 | |
| VCDBackbone=InternVL2.5-8B, Model Category=Open-Source2026.06 | 23.3 | |
| Multi-ResBackbone=InternVL2.5-8B, Model Category=Open-Source2026.06 | 22.7 | |
| Base ModelBackbone=InternVL2.5-8B, Model Category=Open-Source2026.06 | 21.3 |