Qwen 3 VL 235B A22B Instruct: benchmark results
Alibaba Qwen 3 VL 235B A22B vision-language instruction-tuned checkpoint. Provider: Alibaba. Released 2025-09-22. Access: Open.
Unified ELO 1606 ± 1, rank #245 of 1392 rated models, from 154 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EmpathyBench | 51.3 | Average Score (%) | 100 |
| Konkur 1404 - Humanities | 52.15 | Accuracy (%, text-only) | 100 |
| Konkur 1404 - Mathematics | 60 | Accuracy (%, text-only) | 100 |
| Konkur 1404 - Overall | 50.54 | Accuracy (%, text-only) | 100 |
| LLM Stats (CharadesSTA) | 64.8 | Score (%) | 100 |
| LLM Stats (DocVQAtest) | 97.1 | Score (%) | 100 |
| Math-VR | 65 | Overall Answer Correctness (self-reported) | 96.6 |
| FlagEval EmbodiedVerse - CV-Bench (test) | 88.78 | Score | 95.7 |
| FlagEval EmbodiedVerse - OmniSpatial | 53.85 | Score | 95.7 |
| FlagEval EmbodiedVerse - Where2Place | 58.35 | Score | 95.7 |
| LLM Stats (CC-OCR) | 82.2 | Score (%) | 94.1 |
| CFMME | 63.31 | Average (self-reported) | 93.3 |
Interactive version: theaggregate.ai/model?slug=qwen-3-vl-235b-a22b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-05.