AI Model Radar
中文 EN

Qwen3 VL 235B A22B

Alibaba / Qwen
Model overview

Qwen3-VL-235B-A22B-Instruct

Qwen's current public Qwen3 VL 235B A22B flagship multimodal open-weight family.

multimodalvision-languagereasoningcodingagentlong contextchatdocument understandingcode generationlocal inference
CompareCompanyRankingsPublic signals
Representative version

Qwen3-VL-235B-A22B-Instruct

CompanyAlibaba / Qwen
Representative basisVariant referenced by model provenance
2026-07-15
ReleaseNot disclosed
Release typeModel release
Parameters235B / 22B active
ArchitectureNot disclosed
Context256,000
Input modalitiestext / image / video
Output modalitiestext
Tool use / function callingConfirmed: Tool use
Structured output / reasoningNot shown: Reasoning
Vision / audio / videoConfirmed: Vision, Video; Not shown: Audio
Open weights / LicenseWeights available / Apache-2.0
License sourcealibaba-qwen-qwen3-vl-235b-a22b-instruct-current · 2026-07-15
Vendor APINot disclosed
API source pending verification
OpenRouterOpenRouter listed
PricingOpenRouter · 2026-07-15: $0.21 / $1.9
input / output, per 1M tokens
BenchmarkNo reliable public source
Public signals1 runtime / access record / 1 community discussion / 2 activity snapshots
Use casesassistant chat / document understanding / code generation / local inference
Claim evidenceRelease / model specs / weightsQwen/Qwen3-VL-235B-A22B-Instruct · Hugging FaceQwen3-VL-235B-A22B-Instruct
OpenRouter

Pricing / context / availability

OpenRouterOpenRouter listed
OpenRouter listingqwen/qwen3-vl-235b-a22b-instruct
Mapped versionQwen3-VL-235B-A22B-Instruct
alibaba-qwen-qwen3-vl-235b-a22b-instruct-current
PricingOpenRouter · 2026-07-15: $0.21 / $1.9
input / output, per 1M tokens
Context131072
Speed/LatencyNo reliable public source
Snapshot time2026-07-15T20:11:41.551Z
SourceView source
Public activity counters

Cumulative Hugging Face model-card downloads / likes / spaces

PlatformEntryDownloadsLikesSpacesTimeSource
huggingfaceQwen/Qwen3-VL-235B-A22B-Instruct1,720,887407902026-07-15T20:11:41.551ZView source
huggingfaceQwen/Qwen3-VL-235B-A22B-Thinking13,381400652026-07-15T20:11:41.551ZView source
Operations

Runtime / access signals

TimeSignalOriginal textSource
01/29/2026, 02:32 PM UTCQwen3-VL-235B-A22B-Instruct-AWQ returns chinese charactersUpon inferencing the model, the model produces random chinese characters from the document's image, even though the whole document is english....Source
Vertical comparison

Current available versions

VersionReleaseParametersContextInputOutputOpen weights / LicenseVendor APISource tier
Qwen3-VL-235B-A22B-InstructNot disclosed235B / 22B active256,000text / image / videotextWeights available / Apache-2.0Not disclosedP0
Qwen3-VL-235B-A22B-ThinkingNot disclosed235B / 22B active256,000text / image / videotextWeights available / Apache-2.0Not disclosedP0
Evidence gaps

Public display boundary

Release date not disclosedNo reliable benchmark sourceNo linkable real reviewSpeed/latency not disclosed