Saltar al contenido principal

Qwen3 VL 30B A3B (Reasoning)

AlibabaQwenOpen WeightApache 2.0 · Uso Comercial

Descripción

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.

Fecha de lanzamiento
2025-10-03
Parámetros
31.0B
Longitud del contexto
131K
Modalidades
image, text, video

Radar de capacidades

33
general
61
coding
81
reasoning
44
scienceest.
70
agents
100
multimodal

Science usa un proxy de razonamiento cuando los benchmarks científicos dedicados no están disponibles.

Rankings

Dominio#PosiciónPuntuaciónFuente
Capacidad agéntica69
54.0
LS
Ranking de codificación211
46.0
AA
Ranking general276
39.0
AA
Razonamiento matemático73
83.0
AA
Ranking multimodal42
77.0
LS
Razonamiento85
53.0
LS
Ciencia247
45.0
AA

Puntuaciones de benchmarks (LLM Stats)

3d

BLINK65.4%Aut.

Agents

BFCL-v368.6%Aut.
OSWorld30.6%Aut.

Biology

GPQA74.4%Aut.

Chemistry

SuperGPQA56.4%Aut.

Communication

MM-MT-Bench7.90 / 100Aut.
WritingBench85.2%Aut.
Multi-IF73.0%Aut.

Creativity

Creative Writing v382.5%Aut.
Arena-Hard v256.7%Aut.

Factuality

SimpleQA23.9%Aut.

Finance

MMLU87.6%Aut.
MMLU-Pro80.5%Aut.
MMLU-ProX76.1%Aut.

General

MMLU-Redux90.9%Aut.
IFEval81.7%Aut.
MLVU-M78.9%Aut.
MMMU (val)76.0%Aut.
MMStar75.5%Aut.
Include74.5%Aut.
LiveBench 2024112572.1%Aut.
LiveCodeBench v664.2%Aut.
MMMU-Pro63.0%Aut.

Grounding

ScreenSpot94.7%Aut.
ScreenSpot Pro57.3%Aut.

Healthcare

VideoMMMU75.0%Aut.

Image To Text

OCRBench83.9%Aut.
OCRBench-V2 (en)62.6%Aut.
OCRBench-V2 (zh)60.4%Aut.

Language

CharadesSTA62.7%Aut.

Long Context

LVBench59.2%Aut.

Math

AIME 202583.1%Aut.
MathVista-Mini81.9%Aut.
HMMT2567.6%Aut.
MathVision65.7%Aut.
PolyMATH51.7%Aut.

Multimodal

DocVQAtest95.0%Aut.
MMBench-V1.188.9%Aut.
CharXiv-D86.9%Aut.
AI2D86.9%Aut.
InfoVQAtest86.0%Aut.
CC-OCR77.8%Aut.
MuirBench77.6%Aut.
Video-MME73.3%Aut.
MVBench72.0%Aut.
CharXiv-R56.6%Aut.

Reasoning

Hallusion Bench66.0%Aut.
ERQA45.3%Aut.

Spatial Reasoning

RealWorldQA77.4%Aut.

Vision

ODinW42.3%Aut.

Índices de evaluación AA

Math Index
82.3
Intelligence Index
13.3
Aime 25
0.8
Mmlu Pro
0.8
Gpqa
0.7
Livecodebench
0.7
Ifbench
0.5
Lcr
0.4
Scicode
0.3
Tau2
0.2
Hle
0.1
Terminalbench Hard
0.1

Puntuaciones por categoría LLM Stats

Communication
3
Multimodal
100
Instruction Following
80
Language
80
Legal
80
Structured Output
80
Finance
80
Grounding
80
Healthcare
80
Text-to-image
80
Image To Text
70
Math
70
Physics
70
Reasoning
70
Spatial Reasoning
70
General
70
3d
70
Biology
70
Chemistry
70
Creativity
70
Tool Calling
70
Video
70
Vision
70
Writing
70
Long Context
60
Economics
60
Agents
50
Factuality
20

Precios

Precio de entrada$0.2 / 1M tokens
Precio de salida$0.75 / 1M tokens
Precio mixto (3:1)$0.338 / 1M tokens

Velocidad

Tokens/seg122.7
Retraso del primer token1.01s
Tiempo hasta la respuesta17.31s

Ranking de Precios por Proveedor

Ranking de Precios por Proveedor

2 proveedores

Más barato: Alibaba (China)Más caro: Alibaba
ProveedorEntradaSalida
1Alibaba (China)Más barato
$0.108
$0.431
2AlibabaPRINCIPAL
$0.2
$0.75

Comparar precios entre diferentes proveedores de API para este modelo.

Fuentes externas