all the models — AI benchmark observatory
← Models

Qwen2.5 VL 32B Instruct

Alibaba Cloud / Qwen Team · open weight · qwen2.5-vl-32b

Qwen2.5-VL is a vision-language model from the Qwen family. Key enhancements include visual understanding (objects, text, charts, layouts), visual agent capabilities (tool use, computer/phone control), long video comprehension with event pinpointing, visual localization (bounding boxes/points), and structured output generation.

DocVQAAndroid ContHumanEvalScreenSpotAITZ_EMMATH

Benchmark scores

BenchmarkScore
DocVQA0.95
Android Control Low_EM0.93
HumanEval0.92
ScreenSpot0.89
AITZ_EM0.83
MATH0.82
MMLU0.78
MMMU0.70
MMLU-Pro0.69
MMMU-Pro0.49
GPQA0.46

Pricing

  • No provider pricing.

AA metrics

No Artificial Analysis link yet.

Arena Elo

  • No Arena snapshot linked.