| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| WildVision | Win Rate80.6 | 25 | 10d ago | ||
| MME-RealWorld EN | InternVL2.5-38B | Score64 | 20 | 1mo ago | |
| R-Bench | Phi-3.5-Vision-4B | Distance Error55.5 | 19 | 1mo ago | |
| V* Bench | DeepEyes | Accuracy85.6 | 18 | 1mo ago | |
| MME RealWorld | Gemini 2.5 Pro | Score71.6 | 11 | 1mo ago | |
| HRBench 4K | Gemini 2.5 Pro | Score86.9 | 10 | 1mo ago | |
| RealWorldQA | Defender Iter. 3 | Score70.07 | 9 | 1mo ago | |
| HRBench 8K | DeepEyesV2 | Accuracy73.8 | 8 | 1mo ago | |
| HRBench (4K) | DeepEyesV2 | Accuracy77.9 | 8 | 1mo ago | |
| Tree Bench | Qwen2.5-VL | Accuracy42.5 | 7 | 1mo ago | |
| Real-world Understanding (RealWorldQA, MME-RW, R-Bench) | jina-vlm | RealWorld QA Score68.2 | 5 | 1mo ago |
