| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| MMMU Pro | Qwen3-VL-32B-Instruct | Accuracy56.9 | 31 | 1mo ago | |
| BLINK | Accuracy73.5 | 29 | 25d ago | ||
| ChartQA | Qwen3-VL-8B | Accuracy88.1 | 17 | 25d ago | |
| MM* | Accuracy73.6 | 17 | 25d ago | ||
| General Multimodal Evaluation Suite (MMMU, MMBench, MME, ChartQA, AI2D, HallBench) | MMMU (Val)72.6 | 14 | 5mo ago | ||
| Aggregated Benchmarks | Qwen2.5-VL-3B | Average Score71 | 13 | 2mo ago | |
| Combined 9 Benchmarks | Average Accuracy100 | 13 | 4mo ago | ||
| MME | Normalized Score66.63 | 12 | 2mo ago | ||
| General Benchmarks | JARVIS | Average Score74 | 12 | 5mo ago | |
| MUIR | Accuracy77.6 | 10 | 4mo ago | ||
| Vision-Language Benchmark Suite | SmoothSMoE (k=2.5) | Average Score55.07 | 3 | 1mo ago |
