| Benchmark | Model | Rank | Results |
|---|---|---|---|
| code-generation-on-humaneval | Qwen2.5-VL-72B | #18 | Pass@1: 87.8 |
| video-question-answering-on-implicitqa | Qwen 2.5 VL - 7B | #4 | Average Accuracy: 42.8Macro Average Accuracy: 46.1 |
| visual-question-answering-vqa-on-vlm2-bench | Qwen2.5-VL-7B | #1 | GC-mat: 35.91GC-trk: 43.38OC-cpr: 71.39OC-cnt: 41.72… |