| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| RealWorldQA | GPT-4o | Accuracy75.4 | 68 | 1mo ago | |
| V* Bench (Overall) | Qwen3-VL-4B + SD-RPN | Attribute Score96.5 | 24 | 3mo ago | |
| MMVP | GPT-4o | Accuracy86.3 | 21 | 4mo ago | |
| CV Bench | Long Grounded Thoughts (SFT + GRPO) | Accuracy83.8 | 12 | 5mo ago | |
| MuirBench | GPT-4o | Accuracy68 | 11 | 4mo ago | |
| MMStar-V | Long Grounded Thoughts (SFT + GRPO) | Accuracy68.4 | 10 | 5mo ago | |
| Relative Reflect | LVR-7B | Accuracy41.8 | 4 | 1mo ago | |
| IQ (test) | GPT-4o | Accuracy30 | 4 | 1mo ago |