跳转到主要内容

Qwen3 VL 8B (Reasoning)

AlibabaQwen开源权重Apache 2.0 · 商用许可

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.

发布日期
2025-10-14
参数规模
9.0B
上下文长度
262K
支持模态
image, text

能力雷达图

80
general
0
coding
70
reasoning
60
science估算
60
agents
100
multimodal

缺少专门科学评测时,Science 由 LLM Stats 科学得分或推理能力估算。

排行榜排名

领域#排名分数来源
多模态榜127
34.0
LS

基准测试分数 (LLM Stats)

(LLM Stats (zeroeval))

Chat

MM-MT-Bench8.00 / 100自报
IFEvalGoogle Research (2023)83.2%自报
Multi-IF75.1%自报

Creativity

Creative Writing v382.4%自报

Factuality

SimpleQA49.6%自报

General

MMLU85.2%自报
MLVU-M75.1%自报
Include69.5%自报
BFCL-v363.0%自报
Arena-Hard v251.1%自报

Language

MMLU-Redux88.8%自报
MMLU-Pro77.3%自报
MMLU-ProX70.7%自报

Math

MathVista-Mini81.4%自报
AIME 202580.3%自报
LiveBench 2024112569.8%自报
MathVision62.7%自报
HMMT2560.6%自报
PolyMATH47.5%自报

Multimodal

VideoMMMU72.8%自报
Video-MME71.8%自报
OSWorld33.9%自报

Reasoning

CharXiv-D85.9%自报
GPQANYU + Cohere + Anthropic (2023)69.9%自报
LiveCodeBench v658.6%自报
CharXiv-R53.0%自报
SuperGPQA51.2%自报

Video

CharadesSTA59.9%自报

Vision

DocVQAtest95.3%自报
ScreenSpot93.6%自报
MMBench-V1.187.5%自报
InfoVQAtest86.0%自报
AI2D84.9%自报
OCRBench81.9%自报
MuirBench76.8%自报
CC-OCR76.3%自报
MMStar75.3%自报
MMMU (val)74.1%自报
RealWorldQA73.5%自报
MVBench69.0%自报
BLINK68.7%自报
Hallusion Bench65.4%自报
OCRBench-V2 (en)63.9%自报
MMMU-Pro60.4%自报
OCRBench-V2 (zh)59.2%自报
LVBench55.8%自报
ERQA46.8%自报
ScreenSpot Pro46.6%自报
ODinW39.8%自报

Writing

WritingBench85.5%自报

AA 评测指数

(Artificial Analysis)

暂无 AA 评测数据

LLM Stats 分类评分

(LLM Stats (zeroeval))
Chat
3
Communication
3
Multimodal
100
Instruction Following
80
Language
80
Structured Output
80
General
80
Text-to-image
80
Image To Text
70
Legal
70
Math
70
Reasoning
70
Finance
70
Grounding
70
Healthcare
70
3d
70
Biology
70
Creativity
70
Vision
70
Writing
70
Long Context
60
Physics
60
Spatial Reasoning
60
Chemistry
60
Tool Calling
60
Video
60
Factuality
50
Agents
50
Economics
50

定价

输入价格$0.25 / 1M tokens
输出价格$0.75 / 1M tokens
混合价格(3:1)$0.375 / 1M tokens
缓存读取价格$0.12 / 1M tokens

速度

暂无速度数据

供应商价格排行

供应商价格排行

3 个供应商

最便宜: Alibaba最贵: Kilo Gateway
供应商输入输出
1Alibaba主要
$0.25
$0.75
2OpenRouter
$0.25
$0.75
3Kilo Gateway
$0.25
$0.75

比较该模型在不同 API 供应商之间的定价。

外部链接