| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Counting | CountBench | Accuracy95.1 | 102 | |
| Visual Counting | CountBench | Score90.2 | 36 | |
| Visual Counting | CountBench 13 (test) | Score96.1 | 23 | |
| Counting | CountBench (test) | Accuracy91.5 | 21 | |
| Text-to-Video Generation | CountBench | Counting Accuracy0.591 | 16 | |
| General VQA | CountBench | Accuracy97.35 | 10 | |
| Open-ended Question Answering | CountBench | Accuracy (%)85.5 | 8 | |
| Pointing & Counting | CountBench | Pass@197.3 | 7 | |
| Object Counting | CountBench | Accuracy71.5 | 7 | |
| Text/Chart/Document Understanding | CountBench | Accuracy86.97 | 3 |