llmboard.aiLeaderboard Center
Overall
Overall RankingOpen Models
Tools
Model DirectoryCompare Models
Capabilities
CodingReasoningMathKnowledgeInstruction Following
Price & Efficiency
Price & ValueCapability vs. PriceRuntime Performance
Modalities
Image GenerationVideo GenerationSpeech ModelsEmbeddings
Core Benchmarks
GPQAMMLU-ProAIME 2025SWE-Bench VerifiedMMLUHumanity's Last ExamLiveCodeBenchMATHHumanEvalMMMU-ProView all benchmarks
Methods
Scoring & Data
393 models668 benchmarks

Leaderboard Center

Overall RankingCodingCore BenchmarksPrice & ValueRuntime Performance

Modalities

All ModelsImage GenerationImage EditingVideo GenerationImage-to-VideoVideo EditingText-to-SpeechSpeech-to-TextEmbeddings

Data & Methods

Scoring MethodAll BenchmarksReasoningMath

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai

Alibaba Cloud / Qwen Team model product

Qwen3 VL 30B A3B Thinking

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.

Updated Aug 12, 2026. Default version: Qwen3 VL 30B A3B Thinking

Compare
LLMBoard score31.8Qwen3 VL 30B A3B Thinking
Coverage100%50 benchmark families
Context window262.1KTokens
Official input priceN/AOfficial price unavailable

On this page

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

Qwen3 VL 30B A3B Thinking Specifications

Technical details for the model's default version.

Version
Qwen3 VL 30B A3B Thinking
Released
Sep 22, 2025
Knowledge cutoff
Unknown
Parameters
31B
Context window
262.1K
Max output
262.1K
Inputs
image, text, video
Outputs
text
Open weights
No
License
Apache 2.0

Qwen3 VL 30B A3B Thinking Capability Profile

This profile uses the model's current scored version. Arena ratings and prices are shown separately.

Qwen3 VL 30B A3B Thinking category scores

Qwen3 VL 30B A3B Thinking Benchmark Results

Benchmark scores for Qwen3 VL 30B A3B Thinking.

50 rows
Columns

Show columns

MMMU (val)76.0%021190.0%CAug 11, 2026
MLVU-M78.9%03871.4%CAug 11, 2026
MuirBench77.6%031180.0%CAug 11, 2026
CharadesSTA62.7%051263.6%CAug 11, 2026
OCRBench-V2 (zh)60.4%051160.0%CAug 11, 2026
InfoVQAtest86.0%061254.5%CAug 11, 2026
ScreenSpot94.7%061666.7%CAug 11, 2026
Hallusion Bench66.0%071660.0%CAug 11, 2026
WritingBench85.2%081550.0%CAug 11, 2026
CharXiv-D86.9%091646.7%CAug 11, 2026
LiveBench 2024112572.1%091438.5%CAug 11, 2026
MMBench-V1.188.9%091852.9%CAug 11, 2026
OCRBench-V2 (en)62.6%091227.3%CAug 11, 2026
DocVQAtest95.0%101110.0%CAug 11, 2026
MMStar75.5%102257.1%CAug 11, 2026
MVBench72.0%101743.8%CAug 11, 2026
BLINK65.4%111316.7%CAug 11, 2026
Creative Writing v382.5%111316.7%CAug 11, 2026
MathVista-Mini81.9%112354.5%CAug 11, 2026
MM-MT-Bench7.9 points121731.3%CAug 11, 2026
Multi-IF73.0%122042.1%CAug 11, 2026
PolyMATH51.7%122350.0%CAug 11, 2026
Arena-Hard v256.7%131620.0%CAug 11, 2026
BFCL-v368.6%131933.3%CAug 11, 2026
ODinW42.3%131620.0%CAug 11, 2026
Video-MME73.3%131725.0%CAug 11, 2026
CC-OCR77.8%141823.5%CAug 11, 2026
LVBench59.2%152439.1%CAug 11, 2026
OSWorld30.6%162021.1%CAug 11, 2026
RealWorldQA77.4%162640.0%CAug 11, 2026
HMMT2567.6%172533.3%CAug 11, 2026
Include74.5%173146.7%CAug 11, 2026
OCRBench83.9%172223.8%CAug 11, 2026
AI2D86.9%183245.2%CAug 11, 2026
MMLU-ProX76.1%183245.2%CAug 11, 2026
ScreenSpot Pro57.3%182529.2%CAug 11, 2026
MathVision65.7%193241.9%CAug 11, 2026
VideoMMMU75.0%192628.0%CAug 11, 2026
ERQA45.3%202313.6%CAug 11, 2026
MMLU87.6%2210078.8%CAug 11, 2026
SuperGPQA56.4%233433.3%CAug 11, 2026
MMLU-Redux90.9%244851.1%CAug 11, 2026
SimpleQA23.9%294637.8%CAug 11, 2026
LiveCodeBench v664.2%375330.8%CAug 11, 2026
CharXiv-R56.6%404817.0%CAug 11, 2026
MMMU-Pro63.0%456632.3%CAug 11, 2026
IFEval81.7%506523.4%CAug 11, 2026
MMLU-Pro80.5%5412958.6%CAug 11, 2026
AIME 202583.1%6411444.3%CAug 11, 2026
GPQA74.4%11023453.2%CAug 11, 2026

Qwen3 VL 30B A3B Thinking Arena Results

Preference and agent-evaluation results for the default version.

No Arena results

The default version does not have a matching Arena result yet.

Qwen3 VL 30B A3B Thinking Runtime Performance

Provider-specific output speed and catalog latency for Qwen3 VL 30B A3B Thinking. Runtime does not affect the capability score.

No runtime data

No provider-specific speed or latency record is linked to the default version yet.

Browse runtime rankings

Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.

Qwen3 VL 30B A3B Thinking Pricing

Official vendor API pricing appears first, followed by individual provider offers.

Official API
N/A
Official provider
N/A
Lowest third-party
From $0.20 input, $2.4 output per 1M via OpenRouter
Tracked offerings
2
1 rows
Columns

Show columns

OpenRouterqwen/qwen3-vl-30b-a3b-thinkingglobal$0.20$2.4262.1KAug 11, 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 30B A3B Thinking Versions

Available versions of this model. The score column identifies the version used in the overall ranking.

1 rows
Columns

Show columns

Qwen3 VL 30B A3B ThinkingSep 22, 202531.831B262.1K262.1KNoApache 2.0

What is Qwen3 VL 30B A3B Thinking?

Key information about Qwen3 VL 30B A3B Thinking and its available data.

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.

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.

Data as of 2026-08-11.

Qwen3 VL 30B A3B Thinking vs nearby models

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

Qwen3 VL 30B A3B ThinkingvsClaude Sonnet 3.5Qwen3 VL 30B A3B ThinkingvsQwen3 30B A3BQwen3 VL 30B A3B ThinkingvsGPT-4.1Qwen3 VL 30B A3B ThinkingvsQwen3 32BQwen3 VL 30B A3B ThinkingvsGPT-5-nanoQwen3 VL 30B A3B ThinkingvsLlama 3.1 Nemotron Ultra 253B

Models similar to Qwen3 VL 30B A3B Thinking

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

#156+0.3
AC

Qwen3 30B A3B

Alibaba Cloud / Qwen Team

32.1 LLMBoard

DetailsCompare
#159-0.6
AC

Qwen3 32B

Alibaba Cloud / Qwen Team

31.2 LLMBoard

DetailsCompare
#153+0.7
AC

Qwen3.5 4B

Alibaba Cloud / Qwen Team

32.4 LLMBoard

DetailsCompare
#165-3.0
AC

Qwen3 VL 30B A3B

Alibaba Cloud / Qwen Team

28.8 LLMBoard

DetailsCompare
#166-3.3
AC

Qwen3 VL 8B Thinking

Alibaba Cloud / Qwen Team

28.5 LLMBoard

DetailsCompare
#146+3.5
AC

Qwen3 VL 32B

Alibaba Cloud / Qwen Team

35.2 LLMBoard

DetailsCompare

FAQ

Common questions about Qwen3 VL 30B A3B Thinking.

When was Qwen3 VL 30B A3B Thinking released?

Qwen3 VL 30B A3B Thinking's default version was released on Sep 22, 2025.

How much does Qwen3 VL 30B A3B Thinking cost?

No official standard PAYG price is currently available for Qwen3 VL 30B A3B Thinking. The lowest tracked third-party offer starts at $0.20 input and $2.4 output via OpenRouter.

Who created Qwen3 VL 30B A3B Thinking?

Qwen3 VL 30B A3B Thinking was created by Alibaba Cloud / Qwen Team.

What is the context window for Qwen3 VL 30B A3B Thinking?

The default version has a 262.1K token context window.

Is Qwen3 VL 30B A3B Thinking open weight?

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

How many API providers offer Qwen3 VL 30B A3B Thinking?

2 provider offerings are linked to the default version.

What models should I compare Qwen3 VL 30B A3B Thinking with?

Nearby ranked alternatives include Claude Sonnet 3.5, Qwen3 30B A3B, GPT-4.1.