Skip to main content

Qwen3 VL 235B A22B (Reasoning)

AlibabaQwenOpen WeightApache 2.0 · Commercial OK

Description

Qwen3-VL-235B-A22B-Thinking is the most powerful vision-language model in the Qwen series, featuring 236B parameters with MoE architecture for reasoning-enhanced multimodal understanding. Key capabilities include: Visual Agent (operates PC/mobile GUIs, recognizes elements, invokes tools), Visual Coding (generates Draw.io/HTML/CSS/JS from images/videos), Advanced Spatial Perception (2D grounding and 3D grounding for spatial reasoning and embodied AI), Long Context & Video Understanding (native 256K context expandable to 1M, handles hours-long video with second-level indexing), Enhanced Multimodal Reasoning (excels in STEM/Math with causal analysis), Upgraded Visual Recognition (celebrities, anime, products, landmarks, flora/fauna), and Expanded OCR (32 languages, robust in low light/blur/tilt). Architecture innovations include Interleaved-MRoPE for positional embeddings, DeepStack for multi-level ViT feature fusion, and Text-Timestamp Alignment for precise video temporal modeling.

Release Date
2025-09-23
Parameters
236.0B
Context Length
131K
Modalities
image, text, video

Capability Radar

34
general
65
coding
86
reasoning
57
science
70
agents
100
multimodal

Rankings

Domain#RankScoreSource
Code Ranking283
53.0
AA
General Ranking236
50.0
AA
Multimodal Ranking82
51.0
LS
Science287
48.0
AA

Benchmark Scores (LLM Stats)

(LLM Stats (zeroeval))

Chat

MM-MT-Bench8.50 / 100SR
IFEvalGoogle Research (2023)88.2%SR
Multi-IF79.1%SR

Creativity

Creative Writing v385.7%SR

Factuality

SimpleQA44.4%SR

General

MMLU90.6%SR
Include80.0%SR
SIFO0.77 / 100SR
BFCL-v371.9%SR
SIFO-Multiturn0.71 / 100SR

Language

MMLU-Redux93.7%SR
MMLU-Pro83.8%SR
MMLU-ProX80.6%SR

Math

AIME 202589.7%SR
MathVista-Mini85.8%SR
MathVerse-Mini0.85 / 100SR
LiveBench 2024112579.6%SR
HMMT2577.4%SR
MathVision74.6%SR

Multimodal

VideoMMMU80.0%SR
VideoMME w/o sub.79.0%SR
OSWorld-G0.68 / 100SR
OSWorld38.1%SR

Reasoning

ZebraLogic97.3%SR
LiveCodeBench v670.1%SR
CharXiv-R66.1%SR
SuperGPQA64.3%SR
Humanity's Last Exam13.6%SR

Video

MLVU83.8%SR
CharadesSTA63.5%SR

Vision

DocVQAtest96.5%SR
ScreenSpot95.4%SR
CountBench0.94 / 100SR
Design2Code0.93 / 100SR
MIABench0.93 / 100SR
RefCOCO-avg0.92 / 100SR
MMBench-V1.190.6%SR
InfoVQAtest89.5%SR
AI2D89.2%SR
OCRBench87.5%SR
EmbSpatialBench0.84 / 100SR
CC-OCR81.5%SR
RealWorldQA81.3%SR
MMMUval80.6%SR
MuirBench80.1%SR
MMStar78.7%SR
RoboSpatialHome0.74 / 100SR
Objectron0.71 / 100SR
RefSpatialBench0.70 / 100SR
MMMU-Pro69.3%SR
BLINK67.1%SR
OCRBench-V2 (en)66.8%SR
Hallusion Bench66.7%SR
LVBench63.6%SR
OCRBench-V2 (zh)63.5%SR
ScreenSpot Pro61.8%SR
SimpleVQA0.61 / 100SR
MMLongBench-Doc0.56 / 100SR
ARKitScenes0.54 / 100SR
ERQA52.5%SR
ODinW43.2%SR
SUNRGBD0.35 / 100SR
VisuLogic0.34 / 100SR
ZEROBench-Sub0.28 / 100SR
Hypersim0.11 / 100SR
ZEROBench0.04 / 100SR

Writing

WritingBench86.7%SR

AA Evaluation Indices

(Artificial Analysis)
Aime 25(MAA (Mathematical Association of America))
88.3
Math Index(Artificial Analysis)
88.3
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))
83.6
Gpqa(NYU + Cohere + Anthropic (2023))
77.2
Livecodebench(UC Berkeley + MIT + Cornell (2024))
64.6
Lcr(Artificial Analysis)
63.7
Ifbench(Google Research (2023))
56.5
Tau2(Sierra + U Toronto + Vector Institute (2025))
54.1
Intelligence Index(Artificial Analysis)
13.4
Hle(Center for AI Safety + Scale AI (2025))
11.9
Terminalbench Hard(Stanford × Laude Institute (2026))
11.4

LLM Stats Category Scores

(LLM Stats (zeroeval))
Chat
3
Communication
3
Multimodal
100
Creativity
90
Writing
90
Instruction Following
80
Language
80
Legal
80
Math
80
Structured Output
80
Embodied
80
Finance
80
General
80
Grounding
80
Healthcare
80
Text-to-image
80
Video
80
Image To Text
70
Long Context
70
Reasoning
70
Spatial Reasoning
70
Tool Calling
70
Vision
70
Physics
60
Agents
60
Chemistry
60
Economics
60
Factuality
40
3d
40

Pricing

Input Price$0.4 / 1M tokens
Output Price$4 / 1M tokens
Blended Price (3:1)$1.3 / 1M tokens

Speed

Tokens/sec0.0
Time to First Token0.00s
Time to Answer0.00s

Provider Price Ranking

Provider Price Ranking

11 providers

Cheapest: Venice AIMost Expensive: NovitaAI
ProviderInputOutput
1Venice AICheapest
$0.21
$1.9
2Alibaba (China)
$0.28671
$1.14682
3Merge Gateway
$0.287
$2.867
4AlibabaPRIMARY
$0.4
$4
5OpenRouter
$0.4
$4
6Kilo Gateway
$0.4
$4
7OrcaRouter
$0.4
$4
8Eden AI
$0.4
$4
9NanoGPT
$0.5
$6
10Cortecs
$0.617
$3.119
11NovitaAI
$0.98
$3.95

Compare pricing across different API providers for this model.

External Sources