Qwen 2.5 VL 7B Instruct — benchmark results
Alibaba Qwen 2.5 VL 7B vision-language instruction-tuned checkpoint. Provider: Alibaba. Released 2025-01-28. Access: Open.
Unified ELO 1461 ± 13, rank #950 of 1776 rated models, from 70 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LLM Stats (DocVQA) | 95.7 | Score (%) | 96 |
| LLM Stats (TextVQA) | 84.9 | Score (%) | 92.9 |
| LLM Stats (InfoVQA) | 82.6 | Score (%) | 87.5 |
| EVALITA - MAIA-MC | 81.77 | CPS | 83.3 |
| KOFFVQA - Object Attributes | 78.33 | Score (%) | 77.2 |
| KOFFVQA - Korean OCR | 85 | Score (%) | 75.3 |
| WildRoadBench | 23.6 | AP50 (self-reported) | 75 |
| KOFFVQA - Document Understanding | 74.67 | Score (%) | 69.1 |
| KOFFVQA - Hallucination and Robustness | 75 | Score (%) | 67.9 |
| EVALITA - text-entailment | 78.23 | CPS | 66 |
| LLM Stats (ChartQA) | 87.3 | Score (%) | 65.2 |
| KOFFVQA - Table Understanding | 61.67 | Score (%) | 63 |
Interactive version: theaggregate.ai/model?slug=qwen-2-5-vl-7b-instruct · How the rankings work · Data refreshed daily, snapshot 2026-07-22.