llmboard.ai
Benchmarks
CompareRankings
llmboard.ai
Benchmarks
CompareRankings
HomeModelsQwen3 VL 235B A22B Thinking

Alibaba Cloud / Qwen Team model product

Qwen3 VL 235B A22B Thinking

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.

Updated Aug 10, 2026. Default version: Qwen3 VL 235B A22B Thinking

Compare
LLMBoard score58.5Qwen3 VL 235B A22B Thinking
Coverage60%11 benchmark families
Context window262.1KTokens
Official input priceN/AOfficial price unavailable

On this page

  • Specification
  • Capability
  • Benchmarks
  • Arena
  • Pricing
  • Versions
  • About
  • Compare
  • Similar models
  • FAQ

Model specification

Structured fields from the published default version.

Version
Qwen3 VL 235B A22B Thinking
Released
Sep 22, 2025
Knowledge cutoff
Unknown
Parameters
236B
Context window
262.1K
Max output
262.1K
Inputs
image, text, video
Outputs
text
Open weights
No
License
Apache 2.0

Capability profile

This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.

Qwen3 VL 235B A22B Thinking category scores

Benchmark results

Published benchmark records for the scored version Qwen3 VL 235B A22B Thinking.

67 rows
Columns

Show columns

ARKitScenes0.511100.0%CAug 7, 2026
LiveBench 202411250.8114100.0%CAug 7, 2026
MathVerse-Mini0.811100.0%CAug 7, 2026
MIABench0.911100.0%CAug 7, 2026
MMMUval0.814100.0%CAug 7, 2026
Objectron0.711100.0%CAug 7, 2026
OCRBench-V2 (zh)0.6111100.0%CAug 7, 2026
OSWorld-G0.711100.0%CAug 7, 2026
RoboSpatialHome0.711100.0%CAug 7, 2026
SIFO0.811100.0%CAug 7, 2026
SIFO-Multiturn0.711100.0%CAug 7, 2026
ZebraLogic1.018100.0%CAug 7, 2026
CharadesSTA0.621290.9%CAug 7, 2026
Design2Code0.9220.0%CAug 7, 2026
InfoVQAtest0.921290.9%CAug 7, 2026
MuirBench0.821190.0%CAug 7, 2026
RefSpatialBench0.72680.0%CAug 7, 2026
EmbSpatialBench0.83871.4%CAug 7, 2026
Multi-IF0.832089.5%CAug 7, 2026
RefCOCO-avg0.93766.7%CAug 7, 2026
SUNRGBD0.33433.3%CAug 7, 2026
VisuLogic0.3330.0%CAug 7, 2026
WritingBench0.931585.7%CAug 7, 2026
DocVQAtest1.041170.0%CAug 7, 2026
Hypersim0.1440.0%CAug 7, 2026
OCRBench-V2 (en)0.741272.7%CAug 7, 2026
ScreenSpot1.041680.0%CAug 7, 2026
CC-OCR0.851876.5%CAug 7, 2026
Creative Writing v30.951366.7%CAug 7, 2026
MMLongBench-Doc0.6550.0%CAug 7, 2026
MMLU0.9510096.0%CAug 7, 2026
ZEROBench-Sub0.3550.0%CAug 7, 2026
CountBench0.9660.0%CAug 7, 2026
Hallusion Bench0.761666.7%CAug 7, 2026
MM-MT-Bench8.561768.8%CAug 7, 2026
VideoMME w/o sub.0.861044.4%CAug 7, 2026
BFCL-v30.771966.7%CAug 7, 2026
MMBench-V1.10.971864.7%CAug 7, 2026
MMStar0.872271.4%CAug 7, 2026
MathVista-Mini0.982368.2%CAug 7, 2026
MMLU-Redux0.984885.1%CAug 7, 2026
BLINK0.791333.3%CAug 7, 2026
MLVU0.891011.1%CAug 7, 2026
SimpleVQA0.691333.3%CAug 7, 2026
ZEROBench0.0990.0%CAug 7, 2026
Include0.8103170.0%CAug 7, 2026
MMLU-ProX0.8103271.0%CAug 7, 2026
ODinW0.4101640.0%CAug 7, 2026
OSWorld0.4112047.4%CAug 7, 2026
RealWorldQA0.8112660.0%CAug 7, 2026
LVBench0.6122452.2%CAug 7, 2026
ERQA0.5132345.5%CAug 7, 2026
HMMT250.8132550.0%CAug 7, 2026
OCRBench0.9132242.9%CAug 7, 2026
ScreenSpot Pro0.6132447.8%CAug 7, 2026
SuperGPQA0.6133463.6%CAug 7, 2026
AI2D0.9143258.1%CAug 7, 2026
MathVision0.7143258.1%CAug 7, 2026
VideoMMMU0.8172636.0%CAug 7, 2026
SimpleQA0.4184662.2%CAug 7, 2026
IFEval0.9276559.4%CAug 7, 2026
MMLU-Pro0.82812978.9%CAug 7, 2026
LiveCodeBench v60.7295346.1%CAug 7, 2026
CharXiv-R0.7324732.6%CAug 7, 2026
MMMU-Pro0.7366545.3%CAug 7, 2026
AIME 20250.94911457.5%CAug 7, 2026
Humanity's Last Exam0.1759218.7%CAug 7, 2026

Arena results

Preference and agent-evaluation signals from published Arena datasets.

66 rows
Columns

Show columns

visioncreative writing131229.9219N/AJan 9, 2026
vision style controlcreative writing141217.6219N/AJan 9, 2026
visionhomework621235.0306N/AAug 6, 2026
visionenglish681219.41,096N/AAug 6, 2026
visiondiagram691213.4476N/AAug 6, 2026
visionocr691212.51,347N/AAug 6, 2026
vision style controlhomework691227.3306N/AAug 6, 2026
visionoverall701208.22,347N/AAug 6, 2026
textindustry mathematical731437.3377N/AAug 6, 2026
vision style controlenglish731197.21,096N/AAug 6, 2026
vision style controldiagram751209.6476N/AAug 6, 2026
vision style controlocr751201.21,347N/AAug 6, 2026
vision style controloverall771189.82,347N/AAug 6, 2026
textgerman791422.1170N/AAug 6, 2026
text style controlgerman921420.3170N/AAug 6, 2026
textkorean951363.0211N/AAug 6, 2026
text style controlindustry mathematical981431.7377N/AAug 6, 2026
text style controlkorean1021362.9211N/AAug 6, 2026
textchinese1091451.4296N/AAug 6, 2026
textexpert1111418.8379N/AAug 6, 2026
textmath1121414.0424N/AAug 6, 2026
textcoding1151427.51,627N/AAug 6, 2026
textindustry medicine and healthcare1161416.3433N/AAug 6, 2026
text style controlexpert1191430.3379N/AAug 6, 2026
textenglish1201417.73,882N/AAug 6, 2026
textindustry entertainment and sports and media1201370.61,440N/AAug 6, 2026
textindustry software and it services1201426.92,834N/AAug 6, 2026
textpolish1241378.8389N/AAug 6, 2026
texthard prompts english1251415.52,094N/AAug 6, 2026
textindustry legal and government1251408.1513N/AAug 6, 2026
textspanish1261384.9330N/AAug 6, 2026
text style controlchinese1261438.7296N/AAug 6, 2026
text style controlcoding1261454.61,627N/AAug 6, 2026
texthard prompts1281407.84,039N/AAug 6, 2026
textindustry business and management and financial operations1291391.61,580N/AAug 6, 2026
text style controlmath1291404.0424N/AAug 6, 2026
textoverall1301400.97,934N/AAug 6, 2026
textexclude ties1301379.85,579N/AAug 6, 2026
text style controlindustry software and it services1331436.92,834N/AAug 6, 2026
text style controlpolish1331376.9389N/AAug 6, 2026
textindustry life and physical and social science1341411.81,247N/AAug 6, 2026
textlonger query1341393.61,725N/AAug 6, 2026
text style controlspanish1341384.4330N/AAug 6, 2026
text style controlhard prompts1371418.74,039N/AAug 6, 2026
text style controlnon english1371381.84,052N/AAug 6, 2026
textnon english1381378.84,052N/AAug 6, 2026
text style controlindustry legal and government1381404.3513N/AAug 6, 2026
text style controllonger query1381406.71,725N/AAug 6, 2026
textinstruction following1401374.42,136N/AAug 6, 2026
textmulti turn1401389.71,288N/AAug 6, 2026
text style controlindustry entertainment and sports and media1401361.61,440N/AAug 6, 2026
text style controloverall1421395.47,934N/AAug 6, 2026
textindustry writing and literature and language1431365.21,752N/AAug 6, 2026
text style controlindustry medicine and healthcare1431415.6433N/AAug 6, 2026
text style controlindustry business and management and financial operations1441393.01,580N/AAug 6, 2026
text style controlexclude ties1461372.75,579N/AAug 6, 2026
text style controlinstruction following1461383.12,136N/AAug 6, 2026
text style controlrussian1481377.5481N/AAug 6, 2026
textcreative writing1491344.41,026N/AAug 6, 2026
text style controlindustry life and physical and social science1491406.41,247N/AAug 6, 2026
text style controlenglish1501403.33,882N/AAug 6, 2026
text style controlhard prompts english1501419.62,094N/AAug 6, 2026
textrussian1521360.4481N/AAug 6, 2026
text style controlindustry writing and literature and language1541365.21,752N/AAug 6, 2026
text style controlmulti turn1581387.01,288N/AAug 6, 2026
text style controlcreative writing1671338.51,026N/AAug 6, 2026

Pricing

Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.

Official API
N/A
Official provider
N/A
Lowest third-party
From $0.40 input, $4 output per 1M via OpenRouter
Tracked offerings
4
4 rows
Columns

Show columns

OpenRouterqwen/qwen3-vl-235b-a22b-thinkingglobal$0.40$4131.1KAug 7, 2026
NanoGPTqwen3-vl-235b-a22b-thinkingglobal$0.50$632.8KAug 7, 2026
LLM Gatewayqwen3-vl-235b-a22b-thinkingglobal$0.98$3.95131.1KAug 7, 2026
NovitaAIqwen/qwen3-vl-235b-a22b-thinkingglobal$0.98$3.95131.1KAug 7, 2026

Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.

Qwen3 VL 235B A22B Thinking versions

All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.

1 rows
Columns

Show columns

Qwen3 VL 235B A22B ThinkingSep 22, 202558.5236B262.1K262.1KNoApache 2.0

What is Qwen3 VL 235B A22B Thinking?

A concise description based on the published model registry.

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.

Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.

Data snapshot: 2026-08-07. Editorial model content is not available in the backend.

Qwen3 VL 235B A22B Thinking vs nearby models

Open a comparison with the three ranked models immediately above and below this model.

Qwen3 VL 235B A22B ThinkingvsNemotron 3 SuperQwen3 VL 235B A22B ThinkingvsClaude Opus 3Qwen3 VL 235B A22B ThinkingvsGPT-4.5Qwen3 VL 235B A22B ThinkingvsQwen3.5 27BQwen3 VL 235B A22B ThinkingvsQwen2.5 32BQwen3 VL 235B A22B ThinkingvsGPT-OSS-120B

Models similar to Qwen3 VL 235B A22B Thinking

Recommendations prioritize the same model type and family, then the closest published LLMBoard score.

#58-1.4
AC

Qwen3.5 27B

Alibaba Cloud / Qwen Team

57.1 LLMBoard

DetailsCompare
#59-1.5
AC

Qwen2.5 32B

Alibaba Cloud / Qwen Team

56.9 LLMBoard

DetailsCompare
#50+2.9
AC

Qwen3.5 122B A10B

Alibaba Cloud / Qwen Team

61.4 LLMBoard

DetailsCompare
#70-4.3
AC

Qwen3 Next 80B A3B Thinking

Alibaba Cloud / Qwen Team

54.1 LLMBoard

DetailsCompare
#71-5.2
AC

Qwen3.5 35B A3B

Alibaba Cloud / Qwen Team

53.3 LLMBoard

DetailsCompare
#72-5.4
AC

Qwen2.5 72B

Alibaba Cloud / Qwen Team

53.1 LLMBoard

DetailsCompare

FAQ

Common questions about Qwen3 VL 235B A22B Thinking.

When was Qwen3 VL 235B A22B Thinking released?

Qwen3 VL 235B A22B Thinking's default version was released on Sep 22, 2025.

How much does Qwen3 VL 235B A22B Thinking cost?

No official standard PAYG price is currently available for Qwen3 VL 235B A22B Thinking. The lowest tracked third-party offer starts at $0.40 input and $4 output via OpenRouter.

Who created Qwen3 VL 235B A22B Thinking?

Qwen3 VL 235B A22B Thinking is published under Alibaba Cloud / Qwen Team in the model registry.

What is the context window for Qwen3 VL 235B A22B Thinking?

The default version has a 262.1K token context window.

Is Qwen3 VL 235B A22B Thinking open weight?

No. The default version is not marked as having publicly available weights.

How many API providers offer Qwen3 VL 235B A22B Thinking?

4 published provider offerings are linked to the default version.

What models should I compare Qwen3 VL 235B A22B Thinking with?

Nearby ranked alternatives include Nemotron 3 Super, Claude Opus 3, GPT-4.5.

Rankings

OverallCodingText ArenaPricing

Modalities

Image GenerationVideo GenerationSpeech-to-TextEmbeddings

Benchmarks

All BenchmarksReasoningMathCoding

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai