LLM Reference

Qwen3-VL Models by Alibaba

AlibabaApache 2.0Open source
6 models2025Up to 256k ctxFrom $0.08/1M input

Details

ResearcherAlibaba
LicenseApache 2.0OSI-approved
Commercial useCommercial use: permitted
Models6
Released2025
Max context256k

Capabilities

VisionAll models
MultimodalAll models
Function Calling4 of 6 models
Tool Use4 of 6 models
Structured OutputsAll models

About

Qwen3-VL is a family of 6 AI models by Alibaba, released in 2025.

Current Variants

Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.

6 in view
Qwen3-VL-235B-A22BCurrent

Use when the workload needs 128k context, structured outputs, and multimodal inputs.

2025-12128k contextstructured outputsmultimodal inputs
Qwen3 VL 32B InstructCurrent

Use when the workload needs 128k context, 32B parameters, and tool use.

2025-09128k context32B parameterstool use
Qwen3 VL 8B InstructCurrent

Use when the workload needs 128k context, 8B parameters, and tool use.

2025-09128k context8B parameterstool use
Qwen3 VL 30B A3B InstructCurrent

Use when the workload needs 128k context, 30B parameters, and tool use.

2025-09128k context30B parameterstool use
Qwen3 VL 235B A22B InstructCurrent

Use when the workload needs 256k context, 235B parameters, and tool use.

2025-09256k context235B parameterstool use
Qwen3-VL-30B-A3BCurrent

Use when the workload needs multimodal, 128k context, and 30B parameters.

2025-01multimodal128k context30B parameters

Release Timeline

3 release groups
2025-12
1 current
Qwen3-VL-235B-A22B
128k contextstructured outputsmultimodal inputs
Current
2025-09
4 current
Qwen3 VL 235B A22B Instruct
256k context235B parameterstool use
Current
Qwen3 VL 30B A3B Instruct
128k context30B parameterstool use
Current
Qwen3 VL 32B Instruct
128k context32B parameterstool use
Current
Qwen3 VL 8B Instruct
128k context8B parameterstool use
Current
2025-01
1 current
Qwen3-VL-30B-A3B
multimodal128k context30B parameters
Current

Specifications(6 models)

Qwen3-VL model specifications comparison
ModelReleasedContextParametersVisionMultimodalFn CallingTool UseStructured Outputs
Qwen3-VL-235B-A22B2025-12128k235B (22B active)YesYesNoNoYes
Qwen3 VL 32B Instruct2025-09128k32BYesYesYesYesYes
Qwen3 VL 8B Instruct2025-09128k8BYesYesYesYesYes
Qwen3 VL 30B A3B Instruct2025-09128k30BYesYesYesYesYes
Qwen3 VL 235B A22B Instruct2025-09256k235BYesYesYesYesYes
Qwen3-VL-30B-A3B2025-01128k30BYesYesNoNoYes

Pricing

Qwen3-VL model pricing by provider
ModelProviderInput / 1MOutput / 1MType
Qwen3 VL 8B InstructNovita AI$0.08$0.5Serverless
Qwen3-VL-30B-A3BOpenRouter$0.13$1.56Serverless
Qwen3-VL-30B-A3BFireworks AI$0.15$0.6Serverless
Qwen3 VL 30B A3B InstructNovita AI$0.2$0.7Serverless
Qwen3-VL-235B-A22BOpenRouter$0.26$2.6Serverless
Qwen3 VL 235B A22B InstructNovita AI$0.3$1.5Serverless
Qwen3 VL 235B A22B InstructVercel AI Gateway$0.4$1.6Serverless
Qwen3-VL-235B-A22BAWS Bedrock$0.53$2.66Serverless
Qwen3-VL-235B-A22BNovita AI$0.98$3.95Serverless

Frequently Asked Questions

What is Qwen3-VL used for?
Qwen3-VL is used for multimodal, vision and multimodal work, and agent workflows and tool use. The family description and listed model capabilities point to those workloads as the best fit.
How does Qwen3-VL compare to Tongyi DeepResearch?
Qwen3-VL by Alibaba is strongest where you need multimodal, while Tongyi DeepResearch by Alibaba is the closest related family to check for adjacent model selection. Qwen3-VL has 6 listed variants and reaches up to 256k context, while Tongyi DeepResearch reaches up to 131k context, so compare the specs and pricing tables before choosing a production model.
Which Qwen3-VL model should I use?
For the lowest listed input price, start with Qwen3 VL 8B Instruct through Novita AI at $0.08/1M input tokens. For the most capable/latest local choice, evaluate Qwen3 VL 235B A22B Instruct with 256k context and tool use, function calling, structured outputs, and multimodal inputs.