Skip to main content
AITier

Qwen2.5 VL 32B Instruct

Alibaba Cloud / Qwen TeamQwenOpen WeightApache 2.0 · Commercial OK

Description

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.

Release Date
2025-02-28
Parameters
33.5B
Context Length
—
Modalities
—

Capability Radar

50
general
90
coding
70
reasoning
43
scienceest.
40
agents
70
multimodal

Science uses a reasoning proxy when dedicated science benchmarks are unavailable.

Rankings

Domain#RankScoreSource
Agents & Tools94
33.0
LS
Multimodal Ranking65
66.0
LS

Benchmark Scores (LLM Stats)

Agents

AITZ_EM83.1%SR
AndroidWorld_SR22.0%SR
OSWorld5.9%SR

Biology

GPQA46.0%SR

Code

HumanEval91.5%SR

Finance

MMLU78.4%SR
MMLU-Pro68.8%SR

General

MBPP0.84 / 100SR
MMMU70.0%SR
MMStar69.5%SR
MMMU-Pro49.5%SR

Grounding

ScreenSpot88.5%SR
ScreenSpot Pro39.4%SR

Image To Text

DocVQA94.8%SR
OCRBench-V2 (zh)59.1%SR
OCRBench-V2 (en)57.2%SR

Language

CharadesSTA54.2%SR

Long Context

LVBench49.0%SR

Math

MATH82.2%SR
MathVista-Mini74.7%SR
MathVision38.4%SR

Multimodal

Android Control Low_EM93.3%SR
InfoVQA83.4%SR
VideoMME w sub.77.9%SR
CC-OCR77.1%SR
VideoMME w/o sub.70.5%SR
Android Control High_EM69.6%SR
MMBench-Video1.9%SR

AA Evaluation Indices

No AA evaluation data available

LLM Stats Category Scores

Code
90
Structured Output
80
Text-to-image
80
Finance
70
Healthcare
70
Image To Text
70
Language
70
Legal
70
Math
70
Spatial Reasoning
60
Vision
60
Grounding
60
Multimodal
60
Reasoning
60
Video
50
Biology
50
Chemistry
50
General
50
Long Context
50
Physics
50
Agents
40

Pricing

No pricing data available

Speed

No speed data available

Available Providers

(LS internal units)

No provider data available

External Sources

LLM Stats