Qwen-VL-Max-2025-08-13: benchmark results
Provider: Alibaba.
Aggregate Rating 1688 ± 48, rank #404 of 10916 rated models, from 25 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| UniKIE-Bench (Constrained-Category) - Administrative | 79.96 | Field-level F1 (%) on the 385 administrative documents of Un | 92.9 |
| UniKIE-Bench (Constrained-Category) - Postal Label | 84.85 | Field-level F1 (%) on the 500 postal label documents of UniK | 92.9 |
| UniKIE-Bench (Open-Category) - English Contract | 76.57 | Field-level F1 (%) on the 207 English contract documents of | 92.9 |
| UniKIE-Bench (Constrained-Category) - Catering Services | 84.68 | Field-level F1 (%) on the 212 catering services documents of | 85.7 |
| UniKIE-Bench (Constrained-Category) - Education | 82.46 | Field-level F1 (%) on the 320 education documents of UniKIE- | 85.7 |
| UniKIE-Bench (Open-Category) | 69.95 | Field-level F1 (%) averaged over the eight language and docu | 85.7 |
| UniKIE-Bench (Open-Category) - Chinese Contract | 69.68 | Field-level F1 (%) on the 208 Chinese contract documents of | 85.7 |
| UniKIE-Bench (Open-Category) - Chinese Form | 69.75 | Field-level F1 (%) on the 208 Chinese form documents of UniK | 85.7 |
| UniKIE-Bench (Open-Category) - Chinese Invoice | 66.62 | Field-level F1 (%) on the 207 Chinese invoice documents of U | 85.7 |
| UniKIE-Bench (Open-Category) - Chinese Receipt | 69.94 | Field-level F1 (%) on the 202 Chinese receipt documents of U | 85.7 |
| UniKIE-Bench (Open-Category) - English Form | 65.52 | Field-level F1 (%) on the 210 English form documents of UniK | 85.7 |
| UniKIE-Bench (Open-Category) - English Invoice | 70.8 | Field-level F1 (%) on the 212 English invoice documents of U | 85.7 |
Interactive version: theaggregate.ai/model?slug=qwen-vl-max-2025-08-13 · How It Works · Data refreshed daily, snapshot 2026-11-08.