跳轉到主要內容

Qwen3 VL 32B Instruct

AlibabaQwen開源權重Apache 2.0 · 商用許可

描述

Qwen3-VL is a large multimodal model that unifies vision, language, and reasoning to achieve human-level perception and cognition across text, images, and video. Built on a 235B-parameter architecture, it integrates early joint training of visual and textual modalities for strong language grounding. The model supports up to a 1 million-token context window and excels at visual understanding, spatial reasoning, long video comprehension, and tool-based interaction. It can generate code from images, perform precise 2D/3D object grounding, and operate digital interfaces like a visual agent. The “Instruct” version rivals Gemini 2.5 Pro in perception benchmarks, while the “Thinking” version leads in multimodal reasoning and STEM tasks. With multilingual OCR, creative writing, and fine-grained scene interpretation, Qwen3-VL establishes a new open-source frontier for integrated vision-language intelligence.

發布日期
2025-10-21
參數規模
33.0B
上下文長度
—
支援模態
—

能力雷達圖

90
general
0
coding
70
reasoning
60
science估算
70
agents
90
multimodal

缺少專門科學評測時,Science 由 LLM Stats 科學得分或推理能力估算。

排行榜排名

領域#排名分數來源
多模態榜80
50.0
LS

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Chat

MM-MT-Bench8.40 / 100自報
IFEvalGoogle Research (2023)84.7%自報
Multi-IF72.0%自報

Creativity

Creative Writing v385.6%自報

General

MMLU86.4%自報
MLVU-M82.1%自報
Include74.0%自報
BFCL-v370.2%自報
Arena-Hard v264.7%自報

Language

MMLU-Redux89.8%自報
MMLU-Pro78.6%自報
MMLU-ProX73.4%自報

Math

MathVista-Mini83.8%自報
LiveBench 2024112572.2%自報
AIME 202566.2%自報
MathVision63.4%自報
PolyMATH40.5%自報

Multimodal

OSWorld32.6%自報

Reasoning

CharXiv-D90.5%自報
GPQANYU + Cohere + Anthropic (2023)68.9%自報
CharXiv-R62.8%自報
SuperGPQA54.6%自報
LiveCodeBench v643.8%自報

Video

CharadesSTA61.2%自報

Vision

DocVQAtest96.9%自報
ScreenSpot95.8%自報
OCRBench89.5%自報
AI2D89.5%自報
InfoVQAtest87.0%自報
CC-OCR80.3%自報
RealWorldQA79.0%自報
MMStar77.7%自報
MMMU (val)76.0%自報
MVBench72.8%自報
MuirBench72.8%自報
OCRBench-V2 (en)67.4%自報
BLINK67.3%自報
MMMU-Pro65.3%自報
LVBench63.8%自報
Hallusion Bench63.8%自報
OCRBench-V2 (zh)59.2%自報
ScreenSpot Pro57.9%自報
ERQA48.8%自報
ODinW46.6%自報

Writing

WritingBench82.9%自報

AA 評測指數

(Artificial Analysis)

暫無 AA 評測資料

LLM Stats 分類評分

(LLM Stats (zeroeval))
Chat
3
Communication
3
Multimodal
1
General
90
Instruction Following
80
Language
80
Legal
80
Structured Output
80
Grounding
80
Creativity
80
Text-to-image
80
Writing
80
Image To Text
70
Math
70
Reasoning
70
Spatial Reasoning
70
Finance
70
Healthcare
70
3d
70
Biology
70
Tool Calling
70
Video
70
Vision
70
Long Context
60
Physics
60
Chemistry
60
Agents
50
Economics
50

定價

暫無定價資料

速度

暫無速度資料

供應商價格排行

暫無提供商資料

外部連結