Qwen-VL-Max-Latest: benchmark results
Provider: Alibaba.
Aggregate Rating 1647 ± 40, rank #641 of 10916 rated models, from 25 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| FinEval - Multimodal | 73.8 | Weighted average accuracy (%) | 95 |
| FlagEval EmbodiedVerse - MMSI-Bench | 34.1 | Score | 91.3 |
| FlagEval EmbodiedVerse - SAT | 81.33 | Score | 91.3 |
| FlagEval EmbodiedVerse - VSI-Bench (tiny) | 50.88 | Score | 91.3 |
| FlagEval EmbodiedVerse - EgoPlan-Bench2 | 57.74 | Score | 89.1 |
| FlagEval EmbodiedVerse - OmniSpatial | 50.59 | Score | 82.6 |
| FlagEval EmbodiedVerse - Where2Place | 40.12 | Score | 82.6 |
| FlagEval VQA - Text Understanding (Chinese) | 76.7 | Score | 82.6 |
| FlagEval EmbodiedVerse - RoboSpatial-Home | 58.85 | Score | 78.3 |
| FlagEval EmbodiedVerse - All-Angles Bench | 56.61 | Score | 73.9 |
| FlagEval EmbodiedVerse - CV-Bench (test) | 86.38 | Score | 73.9 |
| FlagEval EmbodiedVerse - ERQA | 48.75 | Score | 69.6 |
Interactive version: theaggregate.ai/model?slug=qwen-vl-max-latest · How It Works · Data refreshed daily, snapshot 2026-11-08.