Qwen-VL-Max-0809: benchmark results
Provider: Alibaba. Released 2024-08-30. Access: API.
Aggregate Rating 1606 ± 36, rank #953 of 10916 rated models, from 433 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM MMT-Bench - Raven Progressive Matrices | 75 | Score (%) | 100 |
| OpenVLM MMT-Bench - Reasoning Segmentation | 94.7 | Score (%) | 100 |
| OpenVLM MMT-Bench - VG | 87.4 | Score (%) | 100 |
| OpenVLM MME - Celebrity | 182.9 | Score | 99.8 |
| OpenVLM MMBench V1.0 EN - FP-C | 89.5 | Accuracy (%) | 99.6 |
| OpenVLM SEEDBench IMG - Instances Counting | 75.8 | Accuracy (%) | 99.6 |
| OpenVLM MMT-Bench - Loc | 77.4 | Score (%) | 99.5 |
| OpenVLM MMT-Bench - Polygon Localization | 75 | Score (%) | 99.3 |
| OpenVLM MME - Commonsense Reasoning | 176.4 | Score | 99.1 |
| OpenVLM MMT-Bench - Building Recognition | 95 | Score (%) | 99 |
| OpenVLM MMT-Bench - Camouflage Object Detection | 85 | Score (%) | 99 |
| OpenVLM MMT-Bench - GUI Install | 70 | Score (%) | 99 |
Interactive version: theaggregate.ai/model?slug=qwen-vl-max-0809 · How It Works · Data refreshed daily, snapshot 2026-11-08.