Qwen-VL Models by Alibaba
Alibaba releases · 51 in the last 12 monthsChangelog →
3 models2023Up to 32K ctxFrom $0.050/1M input
Last refreshed 2026-10-17. Next refresh: weekly.
About
Qwen-VL is Alibaba Cloud’s visual multimodal line in the Qwen (Tongyi Qianwen) series. Models accept image, text, and bounding-box inputs and return text and boxes, with multilingual dialogue, multi-image chat, Chinese open-domain grounding, and fine-grained image recognition. The family includes pretrained Qwen-VL and chat models, plus upgraded Qwen-VL-Plus and Qwen-VL-Max with stronger detailed recognition, text-in-image understanding, visual reasoning, and support for very high-resolution inputs.
Decision facts
- Best fit
- structured outputs
- Capability starting point
- Qwen-VL with 32K context
- Lowest tracked input
- Qwen-VL · $0.050/1M · Replicate API
- Closest related family
- Ovis Image
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
1 in view2 retired
Qwen-VLCurrent
Use when the workload needs 32K context and 7B parameters.
2023-1132K context7B parameters
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Qwen-VL | Use when the workload needs 32K context and 7B parameters. | 2023-11 | 32K context7B parameters | Current |
Release Timeline
1 release group2023-11
1 current · 2 retired
Qwen-VL
Current32K context7B parameters
Qwen-VL-Max
Archived32K contextstructured outputs
Qwen-VL-Plus
Archived32K context72B parametersstructured outputs
Specifications(3 models)
| Model | Released | Context | Parameters | Structured Outputs |
|---|---|---|---|---|
| Qwen-VL | 2023-11 | 32K | 7B | No |
Available From(2 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Qwen-VL | Replicate API | $0.05 | $0.25 | Serverless |
