Model
Qwen2.5-VL-72B-Instruct
Alibaba · open weights
Agentic / Tool Use
Agentic / Tool Use| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| OSWorld | 5.0 | score | qwen2.5-vl-72b-instruct (100 steps) | Independent | medium | os-world.github.io |
| OSWorld | 4.4 | score | qwen2.5-vl-72b-instruct (15 steps) | Independent | medium | os-world.github.io |
Multimodal
Multimodal| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| 4MT-VLM | 31.0 | accuracy | N/A | Independent | medium | arxiv.org |
| EgoGears | 28.5 | accuracy | N/A | Independent | medium | arxiv.org |
| GeoBench | 62.0 | acw_country_% | N/A | Independent | medium | geobench.org |
| SpatialViz-Bench | 33.3 | overall_score | N/A | Independent | medium | arxiv.org |
| Video-MME | 73.5 | overall_(no_subtitles) | N/A | Independent | medium | video-mme.github.io |
| WOVEN | 74.7 | accuracy | N/A | Independent | medium | arxiv.org |
Reproduced by the community
Loading community runs…