Alibaba Cloud / Qwen Team model product
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 17, 2026. Default version: Qwen3 VL 235B A22B Instruct
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Benchmark scores for Qwen3 VL 235B A22B Instruct.
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkCharadesSTA | Score64.8% | Rank01 | Participants12 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkDocVQAtest | Score97.1% | Rank01 | Participants11 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkCC-OCR | Score82.2% | Rank02 | Participants18 | Percentile94.1% | EvidenceC | Evaluated |
| BenchmarkCreative Writing v3 | Score86.5% | Rank02 | Participants13 | Percentile91.7% | EvidenceC | Evaluated |
| BenchmarkMMMUval | Score78.7% | Rank02 | Participants4 | Percentile66.7% | EvidenceC | Evaluated |
| BenchmarkBLINK | Score70.7% | Rank03 | Participants15 | Percentile85.7% | EvidenceC | Evaluated |
| BenchmarkCSimpleQA | Score83.4% | Rank03 | Participants8 | Percentile71.4% | EvidenceC | Evaluated |
| BenchmarkInfoVQAtest | Score89.2% | Rank03 | Participants12 | Percentile81.8% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench v5 | Score61.4% | Rank03 | Participants9 | Percentile75.0% | EvidenceC | Evaluated |
| BenchmarkMultiPL-E | Score86.1% | Rank03 | Participants13 | Percentile83.3% | EvidenceC | Evaluated |
| BenchmarkOCRBench | Score92.0% | Rank03 | Participants24 | Percentile91.3% | EvidenceC | Evaluated |
| BenchmarkOCRBench-V2 (en) | Score67.1% | Rank03 | Participants14 | Percentile84.6% | EvidenceC | Evaluated |
| BenchmarkOCRBench-V2 (zh) | Score61.8% | Rank03 | Participants11 | Percentile80.0% | EvidenceC | Evaluated |
| BenchmarkScreenSpot | Score95.4% | Rank03 | Participants16 | Percentile86.7% | EvidenceC | Evaluated |
| BenchmarkAndroidWorld_SR | Score63.7% | Rank04 | Participants8 | Percentile57.1% | EvidenceC | Evaluated |
| BenchmarkODinW | Score48.6% | Rank04 | Participants16 | Percentile80.0% | EvidenceC | Evaluated |
| BenchmarkArena-Hard v2 | Score77.4% | Rank05 | Participants16 | Percentile73.3% | EvidenceC | Evaluated |
| BenchmarkMM-MT-Bench | Score8.5 points | Rank05 | Participants17 | Percentile75.0% | EvidenceC | Evaluated |
| BenchmarkOSWorld | Score66.7% | Rank05 | Participants20 | Percentile79.0% | EvidenceC | Evaluated |
| BenchmarkVideoMME w/o sub. | Score79.2% | Rank05 | Participants10 | Percentile55.6% | EvidenceC | Evaluated |
| BenchmarkWritingBench | Score85.5% | Rank05 | Participants15 | Percentile71.4% | EvidenceC | Evaluated |
| BenchmarkLiveBench 20241125 | Score74.8% | Rank06 | Participants14 | Percentile61.5% | EvidenceC | Evaluated |
| BenchmarkMuirBench | Score72.8% | Rank06 | Participants12 | Percentile54.5% | EvidenceC | Evaluated |
| BenchmarkMLVU | Score84.3% | Rank08 | Participants10 | Percentile22.2% | EvidenceC | Evaluated |
| BenchmarkMMBench-V1.1 | Score89.9% | Rank08 | Participants20 | Percentile63.2% | EvidenceC | Evaluated |
| BenchmarkMMStar | Score78.4% | Rank08 | Participants24 | Percentile69.6% | EvidenceC | Evaluated |
| BenchmarkMulti-IF | Score76.3% | Rank08 | Participants23 | Percentile68.2% | EvidenceC | Evaluated |
| BenchmarkInclude | Score80.0% | Rank09 | Participants31 | Percentile73.3% | EvidenceC | Evaluated |
| BenchmarkMathVista-Mini | Score84.9% | Rank09 | Participants24 | Percentile65.2% | EvidenceC | Evaluated |
| BenchmarkHallusion Bench | Score63.2% | Rank11 | Participants18 | Percentile41.2% | EvidenceC | Evaluated |
| BenchmarkLVBench | Score67.7% | Rank11 | Participants25 | Percentile58.3% | EvidenceC | Evaluated |
| BenchmarkAI2D | Score89.7% | Rank12 | Participants33 | Percentile65.6% | EvidenceC | Evaluated |
| BenchmarkRealWorldQA | Score79.3% | Rank13 | Participants29 | Percentile57.1% | EvidenceC | Evaluated |
| BenchmarkScreenSpot Pro | Score62.0% | Rank13 | Participants25 | Percentile50.0% | EvidenceC | Evaluated |
| BenchmarkSimpleQA | Score51.9% | Rank13 | Participants47 | Percentile73.9% | EvidenceC | Evaluated |
| BenchmarkBFCL-v3 | Score67.7% | Rank14 | Participants19 | Percentile27.8% | EvidenceC | Evaluated |
| BenchmarkMMLU-ProX | Score77.8% | Rank14 | Participants32 | Percentile58.1% | EvidenceC | Evaluated |
| BenchmarkERQA | Score51.3% | Rank16 | Participants24 | Percentile34.8% | EvidenceC | Evaluated |
| BenchmarkMMLU | Score88.8% | Rank16 | Participants101 | Percentile85.0% | EvidenceC | Evaluated |
| BenchmarkSuperGPQA | Score60.4% | Rank17 | Participants34 | Percentile51.5% | EvidenceC | Evaluated |
| BenchmarkHMMT25 | Score57.4% | Rank19 | Participants25 | Percentile25.0% | EvidenceC | Evaluated |
| BenchmarkMathVision | Score66.5% | Rank19 | Participants33 | Percentile43.8% | EvidenceC | Evaluated |
| BenchmarkMMLU-Redux | Score92.2% | Rank19 | Participants48 | Percentile61.7% | EvidenceC | Evaluated |
| BenchmarkVideoMMMU | Score74.7% | Rank20 | Participants26 | Percentile24.0% | EvidenceC | Evaluated |
| BenchmarkIFEval | Score87.8% | Rank29 | Participants67 | Percentile57.6% | EvidenceC | Evaluated |
| BenchmarkCharXiv-R | Score62.1% | Rank39 | Participants51 | Percentile24.0% | EvidenceC | Evaluated |
| BenchmarkMMMU-Pro | Score68.1% | Rank40 | Participants68 | Percentile41.8% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench v6 | Score54.3% | Rank44 | Participants56 | Percentile21.8% | EvidenceC | Evaluated |
| BenchmarkMMLU-Pro | Score81.8% | Rank44 | Participants134 | Percentile67.7% | EvidenceC | Evaluated |
| BenchmarkAIME 2025 | Score74.7% | Rank79 | Participants115 | Percentile31.6% | EvidenceC | Evaluated |
Preference and agent-evaluation results for the default version.
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenavision | Categoryoverall | Rank50 | Rating / score1246.8 | Votes12,102 | ObservationsN/A | Result date |
| Arenavision style control | Categoryoverall | Rank63 | Rating / score1214.6 | Votes12,102 | ObservationsN/A | Result date |
| Arenatext | Categoryoverall | Rank94 | Rating / score1420.8 | Votes11,560 | ObservationsN/A | Result date |
| Arenatext style control | Categoryoverall | Rank121 | Rating / score1414.5 | Votes11,560 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryoverall | Rank124 | Rating / score1410.0 | Votes3,112 | ObservationsN/A | Result date |
Official vendor API pricing appears first, followed by individual provider offers.
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderLLM Gateway | Provider model IDqwen3-vl-235b-a22b-instruct | Regionglobal | Input / 1M$0.20 | Output / 1M$0.88 | Context262.1K | Updated |
| ProviderOpenRouter | Provider model IDqwen/qwen3-vl-235b-a22b-instruct | Regionglobal | Input / 1M$0.21 | Output / 1M$1.9 | Context262.1K | Updated |
| ProviderTensorX | Provider model IDqwen/qwen3-vl-235b-a22b-instruct | Regionglobal | Input / 1M$0.21 | Output / 1M$1.9 | Context131K | Updated |
| ProviderNanoGPT | Provider model IDqwen/Qwen3-VL-235B-A22B-Instruct | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context128K | Updated |
| ProviderHelicone | Provider model IDqwen3-vl-235b-a22b-instruct | Regionglobal | Input / 1M$0.30 | Output / 1M$1.5 | Context256K | Updated |
| ProviderNovitaAI | Provider model IDqwen/qwen3-vl-235b-a22b-instruct | Regionglobal | Input / 1M$0.30 | Output / 1M$1.5 | Context131.1K | Updated |
| ProviderVercel AI Gateway | Provider model IDalibaba/qwen3-vl-235b-a22b-instruct | Regionglobal | Input / 1M$0.40 | Output / 1M$1.6 | Context131.1K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Provider-specific output speed and catalog latency for Qwen3 VL 235B A22B Instruct. Runtime does not affect the capability score.
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Technical details for the model's default version.
Available versions of this model. The score column identifies the version used in the overall ranking.
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionQwen3 VL 235B A22B Instruct | Released | LLMBoard42.7 | Parameters236B | Context262.1K | Max output262.1K | Open weightsNo | LicenseApache 2.0 |
Open a comparison with the three ranked models immediately above and below this model.
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
Key information about Qwen3 VL 235B A22B 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 2026-08-17.
Common questions about Qwen3 VL 235B A22B.
Qwen3 VL 235B A22B's default version was released on Sep 22, 2025.
No official standard PAYG price is currently available for Qwen3 VL 235B A22B. The lowest tracked third-party offer starts at $0.20 input and $0.88 output via LLM Gateway.
Qwen3 VL 235B A22B was created by Alibaba Cloud / Qwen Team.
The default version has a 262.1K token context window.
No. The default version is not marked as having publicly available weights.
7 provider offerings are linked to the default version.
Nearby ranked alternatives include Nova 2 Omni, GPT-OSS-120B, Step3 VL 10B.