# Qwen3-VL

> Alibaba's open-weight flagship vision-language model that reads documents, watches hours-long video, and acts as a visual agent.

- Maker: Alibaba (Qwen)
- Line: Qwen-VL
- Canonical page: https://ai-tldr.dev/models/qwen3-vl/
- Updated: 2026-06-18

## In short

Qwen3-VL is the latest generation of Alibaba's Qwen-VL line of vision-language models, with the flagship Qwen3-VL-235B-A22B released on September 23, 2025 under an Apache 2.0 license. It is a Mixture-of-Experts model with 235B total parameters and roughly 22B active per token, shipped in both an Instruct edition and a reasoning-focused Thinking edition. It has a 256K-token context window. You can use it through Alibaba Cloud Model Studio and OpenRouter.

## Facts

| Fact | Value | Source |
| --- | --- | --- |
| Context window | 256K (max output 32K) | [1] |
| Input price | $0.20 / 1M tokens (0-32K tier) | [2] |
| Output price | $1.60 / 1M tokens (0-32K tier) | [2] |
| Licence | Apache 2.0 · open weights | [1] |
| Released | 23 September 2025 | [3] |
| Parameters | 235B total / 22B active (flagship MoE); also dense 2B/4B/8B/32B and MoE 30B-A3B | [1] |
| Knowledge cutoff | Not publicly disclosed | [1] |

## Sources

1. [Qwen3-VL-235B-A22B-Instruct model card (Hugging Face)](https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct)
2. [Alibaba Cloud Model Studio (API docs & pricing)](https://www.alibabacloud.com/help/en/model-studio/models)
3. [Qwen3-VL GitHub](https://github.com/QwenLM/Qwen3-VL)

## Overview

Qwen3-VL is the latest generation of Alibaba's Qwen-VL line of vision-language models, with the flagship Qwen3-VL-235B-A22B released on September 23, 2025 under an Apache 2.0 license. It is a Mixture-of-Experts model with 235B total parameters and roughly 22B active per token, shipped in both an Instruct edition and a reasoning-focused Thinking edition. The full family also includes dense 2B, 4B, 8B, and 32B models plus a smaller 30B-A3B MoE, so the same architecture scales from edge devices to cloud servers.

Qwen3-VL handles text, images, video, and documents/PDFs in a single interleaved context of 256K tokens natively, expandable to 1 million tokens via YaRN. That window lets it read entire books and reason over very long videos; Alibaba reports the model can localize events across hours-long footage thanks to its Text-Timestamp Alignment design. OCR has been expanded to 32 languages (up from 19 in the previous generation), with better handling of low light, blur, tilt, rare characters, and long-document structure.

Beyond perception, Qwen3-VL is built to act. It works as a visual agent that recognizes GUI elements on PC and mobile screens, understands their function, invokes tools, and completes multi-step tasks. Alibaba states the Instruct flagship matches or exceeds Gemini 2.5 Pro on major visual-perception benchmarks, while the Thinking edition targets state-of-the-art multimodal reasoning. Open weights are on Hugging Face, and a hosted API (qwen3-vl-plus / qwen3-vl-flash) is available through Alibaba Cloud Model Studio.

## Benchmarks

- MathVista: 85.8% ([source](https://the-decoder.com/qwen3-vl-can-scan-two-hour-videos-and-pinpoint-nearly-every-detail/))
- MathVision: 74.6% ([source](https://the-decoder.com/qwen3-vl-can-scan-two-hour-videos-and-pinpoint-nearly-every-detail/))
- DocVQA: 96.5% ([source](https://the-decoder.com/qwen3-vl-can-scan-two-hour-videos-and-pinpoint-nearly-every-detail/))
- OCRBench: 875% ([source](https://the-decoder.com/qwen3-vl-can-scan-two-hour-videos-and-pinpoint-nearly-every-detail/))
- MMMU-Pro: 69.3% ([source](https://the-decoder.com/qwen3-vl-can-scan-two-hour-videos-and-pinpoint-nearly-every-detail/))
- ScreenSpot Pro: 61.8% ([source](https://the-decoder.com/qwen3-vl-can-scan-two-hour-videos-and-pinpoint-nearly-every-detail/))
- AndroidWorld: 63.7% ([source](https://the-decoder.com/qwen3-vl-can-scan-two-hour-videos-and-pinpoint-nearly-every-detail/))

## Pricing

- Input: $0.20 / 1M tokens (0-32K tier) per 1M tokens
- Output: $1.60 / 1M tokens (0-32K tier) per 1M tokens

Hosted qwen3-vl-plus on Alibaba Cloud Model Studio (international); tiered higher for longer context (up to $0.60 input / $4.80 output at 128K-256K). qwen3-vl-flash is cheaper. Open weights can be self-hosted for free under Apache 2.0; third-party hosts such as OpenRouter list the 235B-A22B Instruct model around $0.20 input / $0.88 output per 1M tokens.

Source: https://www.alibabacloud.com/help/en/model-studio/models

## Strengths

- Open weights under a permissive Apache 2.0 license, free to download, fine-tune, and self-host
- 256K native context expandable to 1M tokens for long documents and hours-long video
- Strong visual math and reasoning (MathVista 85.8, MathVision 74.6)
- High-accuracy document understanding and OCR across 32 languages (DocVQA 96.5, OCRBench 875)
- Visual-agent capabilities: operates PC and mobile GUIs, recognizes elements, and invokes tools
- Family scales from a 2B edge model to a 235B MoE flagship, with Instruct and Thinking editions

## Best for

- Document, invoice, and form parsing with multilingual OCR
- Long-video understanding, search, and timestamp-grounded event localization
- GUI automation and visual agents that operate apps on PC and mobile
- Multimodal STEM and visual-math problem solving
- Chart, diagram, and screenshot question answering
- Self-hosted multimodal applications where open weights and data control matter

## How to access

- [Alibaba Cloud Model Studio](https://www.alibabacloud.com/help/en/model-studio/models) — model id `qwen3-vl-plus`
- [OpenRouter](https://openrouter.ai/qwen/qwen3-vl-235b-a22b-instruct) — model id `qwen/qwen3-vl-235b-a22b-instruct`

## Qwen-VL versions

- Qwen3-VL (23 September 2025) — this page
- [Qwen2.5-VL](https://ai-tldr.dev/models/qwen2-5-vl/) (January 2025)
- [Qwen2-VL](https://ai-tldr.dev/models/qwen2-vl/) (December 2024)

## FAQ

### Is Qwen3-VL open source?

Yes. The Qwen3-VL weights, including the flagship 235B-A22B model and the dense 2B/4B/8B/32B variants, are released under the Apache 2.0 license, so you can download, fine-tune, and self-host them, including for commercial use.

### What is the context window of Qwen3-VL?

It supports a 256K-token context window natively, covering interleaved text, images, and video, and can be extended to 1 million tokens using YaRN. That is large enough to process entire books or hours-long videos in a single pass.

### How big is the flagship Qwen3-VL model?

The flagship Qwen3-VL-235B-A22B is a Mixture-of-Experts model with 235B total parameters and about 22B active per token. It comes in an Instruct edition and a reasoning-focused Thinking edition; the full release weighs around 471 GB.

### Can Qwen3-VL act as an agent or do OCR?

Yes. Qwen3-VL works as a visual agent that recognizes GUI elements on PC and mobile screens and invokes tools to complete tasks, and its OCR now covers 32 languages with strong document-parsing results (DocVQA 96.5, OCRBench 875).

## Links

- [Qwen3-VL GitHub](https://github.com/QwenLM/Qwen3-VL)
- [Qwen3-VL-235B-A22B-Instruct model card (Hugging Face)](https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct)
- [Qwen3-VL Technical Report (arXiv)](https://arxiv.org/abs/2511.21631)
- [Alibaba Cloud Model Studio (API docs & pricing)](https://www.alibabacloud.com/help/en/model-studio/models)
- [OpenRouter: Qwen3-VL-235B-A22B-Instruct](https://openrouter.ai/qwen/qwen3-vl-235b-a22b-instruct)
- [Qwen3-VL release coverage (The Decoder)](https://the-decoder.com/qwen3-vl-can-scan-two-hour-videos-and-pinpoint-nearly-every-detail/)
- [Simon Willison: Qwen3-VL launch notes](https://simonwillison.net/2025/Sep/23/qwen3-vl/)
- [Qwen3-VL Demo (Hugging Face Space)](https://huggingface.co/spaces/Qwen/Qwen3-VL-Demo)

Source: AI/TLDR LLM registry, https://ai-tldr.dev/models/qwen3-vl/