# Qwen3 VL 235B Instruct (image)

Demanding image understanding and visual agents. Qwen's flagship vision-language MoE (22B active), for visual reasoning, OCR and GUI agents. Served on GraphN managed inference behind an OpenAI-compatible API as `qwen3-vl-235b`.

## Overview

Qwen3-VL-235B-A22B-Instruct is the largest instruct model in Qwen's Qwen3-VL vision-language series, a mixture of experts with 235B total and 22B active parameters. Qwen highlights operating PC and mobile GUIs as a visual agent, generating HTML, CSS and JavaScript from images, stronger spatial perception, OCR across 32 languages, and text understanding on par with its pure language models.

## Strengths

- The most capable vision model in the catalogue
- OCR across 32 languages, including hard scans
- Reads and acts on GUI screenshots

## Use cases

- Document and chart understanding
- Screenshot-driven GUI agents
- Turning mockups into front-end code

## Good to know

Only image input is supported on this alias; video input is not.

## Specifications

- **Type**: Chat
- **Context**: 256K context
- **Parameters**: 235B total, 22B active (MoE)
- **License**: Apache-2.0
- **Family**: Qwen3-VL
- **Weights**: Qwen/Qwen3-VL-235B-A22B-Instruct
- **Pricing**: Input $0.53/1M · Output $2.66/1M

## API access

### curl

```bash
curl https://model.graphn.ai/v1/chat/completions \
  -H "Authorization: Bearer $GRAPHN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3-vl-235b",
    "messages": [{"role": "user", "content": "Hello"}]
  }'
```

### CLI

```bash
graphn model chat qwen3-vl-235b -m "What is GraphN?"
```

### Python

```python
@function
async def ask(question: str) -> str:
    return await chat.complete(question, model="qwen3-vl-235b")
```

[Start free](https://graphn.ai/signup?next=%2Finference%2Fqwen3-vl-235b)

[Model card](https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct) · [Managed inference](https://graphn.ai/inference) · [Pricing](https://graphn.ai/pricing) · [API reference](https://graphn.ai/api)