メインコンテンツへスキップ

Qwen3 VL 4B Instruct

AlibabaQwenオープンウエイトApache 2.0 · 商用利用可

説明

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.

リリース日
2025-10-14
パラメータ
4.0B
コンテキスト長
—
モダリティ
image, text

能力レーダー

23
general
29
coding
37
reasoning
27
science
60
agents
90
multimodal

ランキング

ドメイン#順位スコアソース
コーディングランキング536
16.0
AA
総合ランキング524
26.0
AA
マルチモーダルランキング138
26.0
LS
科学580
18.0
AA

ベンチマークスコア (LLM Stats)

(LLM Stats (zeroeval))

Chat

MM-MT-Bench7.50 / 100自己申告
IFEvalGoogle Research (2023)82.3%自己申告

Factuality

SimpleQA48.0%自己申告

General

MMLU77.2%自己申告
MLVU-M75.3%自己申告
BFCL-v363.3%自己申告
Include61.4%自己申告

Language

MMLU-Redux81.5%自己申告
MMLU-Pro67.1%自己申告
MMLU-ProX59.4%自己申告

Math

MathVista-Mini73.7%自己申告
LiveBench 2024112560.9%自己申告
MathVision51.6%自己申告
AIME 202546.6%自己申告
HMMT2530.7%自己申告
PolyMATH28.8%自己申告

Multimodal

VideoMMMU56.2%自己申告
OSWorld26.2%自己申告

Reasoning

CharXiv-D76.2%自己申告
SuperGPQA40.3%自己申告
CharXiv-R39.7%自己申告
LiveCodeBench v637.9%自己申告

Video

CharadesSTA55.5%自己申告

Vision

DocVQAtest95.3%自己申告
ScreenSpot94.0%自己申告
OCRBench88.1%自己申告
MMBench-V1.185.1%自己申告
AI2D84.1%自己申告
InfoVQAtest80.3%自己申告
CC-OCR76.2%自己申告
RealWorldQA70.9%自己申告
MMStar69.8%自己申告
MVBench68.9%自己申告
MMMU (val)67.4%自己申告
BLINK65.8%自己申告
MuirBench63.8%自己申告
OCRBench-V2 (en)63.7%自己申告
ScreenSpot Pro59.5%自己申告
OCRBench-V2 (zh)57.6%自己申告
Hallusion Bench57.6%自己申告
LVBench56.2%自己申告
MMMU-Pro53.2%自己申告
ODinW48.2%自己申告
ERQA41.3%自己申告

Writing

WritingBench82.5%自己申告

AA評価指数

(Artificial Analysis)
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))
63.4
Gpqa(NYU + Cohere + Anthropic (2023))
37.1
Math Index(Artificial Analysis)
37.0
Aime 25(MAA (Mathematical Association of America))
37.0
Ifbench(Google Research (2023))
31.8
Livecodebench(UC Berkeley + MIT + Cornell (2024))
29.0
Tau2(Sierra + U Toronto + Vector Institute (2025))
23.4
Lcr(Artificial Analysis)
14.0
Intelligence Index(Artificial Analysis)
5.7
Hle(Center for AI Safety + Scale AI (2025))
3.6
Terminalbench Hard(Stanford × Laude Institute (2026))
0.0

LLM Statsカテゴリスコア

(LLM Stats (zeroeval))
Chat
4
Communication
4
Multimodal
90
Instruction Following
80
General
80
Grounding
80
Creativity
80
Text-to-image
80
Writing
80
Image To Text
70
Language
70
Legal
70
Structured Output
70
3d
70
Long Context
60
Math
60
Reasoning
60
Spatial Reasoning
60
Finance
60
Healthcare
60
Tool Calling
60
Video
60
Vision
60
Factuality
50
Physics
40
Agents
40
Chemistry
40
Economics
40

価格設定

入力価格無料
出力価格無料
混合価格(3:1)無料

速度

トークン/秒0.0
初トークン遅延0.00s
初回答遅延0.00s

プロバイダー価格ランキング

プロバイダーデータがありません

外部リンク