Skip to content
Qwen

Qwen3 VL 235B A22B Instruct

Qwen3 VL 235B A22B Instruct is a large language model made by Qwen, added to the catalog on 23 Sep 2025. It has a context window of 262,144 tokens and costs $0.21 per 1M input tokens and $1.90 per 1M output tokens. It accepts text and images and produces text. Typical uses: enhance document AI workflows, perform multilingual OCR tasks and automate software and UI processes.

Price checked ยท source: OpenRouter

What Qwen3 VL 235B A22B Instruct is

Discover Qwen: Qwen3 VL 235B A22B Instruct by Qwen, released on September 23, 2025. This advanced model combines text and image inputs to produce text outputs, excelling in tasks like VQA, document parsing, and multilingual OCR. With a context window of 262,144 tokens, it offers robust perception and spatial understanding. The model supports complex instructions, aligns text with video, and aids in UI debugging, making it versatile for document AI and vision-language research.

Key features

  • Multimodal text and image processing
  • Supports VQA and document parsing
  • Aligns text with video timelines
  • Robust perception and spatial understanding
  • Aids in UI debugging and automation

Typical uses

  • Enhance document AI workflows
  • Perform multilingual OCR tasks
  • Automate software and UI processes
  • Conduct vision-language research
  • Execute spatial and embodied tasks

Qwen3 VL 235B A22B Instruct pricing

US dollars, list price through OpenRouter.

Checked
Qwen3 VL 235B A22B Instruct price list
ItemStandard
Input, per 1M tokens$0.21
Output, per 1M tokens$1.90
Cached input read, per 1M tokens$0.10

No price change since it was added on 23 Sep 2025.

What typical tasks cost with Qwen3 VL 235B A22B Instruct
What a task costsCost
Answer a short question500 in / 300 out$0.0007
Summarise a 50-page PDF25,000 in / 800 out$0.0068
Review a 2,000-line source file30,000 in / 2,000 out$0.010
Read a 300-page contract set150,000 in / 1,500 out$0.034
Check the provider's pageToken counts are estimates. A maker's direct price and its monthly plans can differ from the per-token price shown here.

Qwen3 VL 235B A22B Instruct specifications

Qwen3 VL 235B A22B Instruct limits and identity
Model idqwen/qwen3-vl-235b-a22b-instruct
Versionqwen3-vl-235b-a22b-instruct
Context window262,144 tokensAbout 196,608 English words
Max output32,768 tokens
Knowledge cutoff2025-03-31
Added to catalog23 Sep 2025
Qwen3 VL 235B A22B Instruct capabilities
Input typestext, image
Output typestext
Tool callingYes
Structured outputsYes
Reasoning effort controlNo
Content moderationNot applied by provider

Sample outputs

The same prompts are given to every model, and the unedited output is shown with the date it was produced. Upend does not publish leaderboard scores it did not run.

Write a short essay

RUN20 Nov 2026
Prompt

Write 150 words on how AI might positively upend work, leisure and creativity

Output

Revolutionize

Draw with code

RUN20 Nov 2026
Prompt

Generate an SVG of a pelican riding a bicycle.

Output
Qwen3 VL 235B A22B Instruct's drawing for the prompt: Generate an SVG of a pelican riding a bicycle.

Qwen3 VL 235B A22B Instruct questions

How much does Qwen3 VL 235B A22B Instruct cost?

Qwen3 VL 235B A22B Instruct costs $0.21 per 1M input tokens and $1.90 per 1M output tokens through OpenRouter, as checked on 22 Nov 2026. Summarising a 50-page PDF costs about $0.0068.

What is the context window of Qwen3 VL 235B A22B Instruct?

262,144 tokens, or roughly 196,608 English words. It can write up to 32,768 tokens in one response.

Can Qwen3 VL 235B A22B Instruct read images and PDFs?

It accepts text and images as input. It produces text, so it cannot generate images.

How do I try Qwen3 VL 235B A22B Instruct?

Use the "Try Qwen3 VL 235B A22B Instruct" button on this page. It opens the chat with Qwen3 VL 235B A22B Instruct selected; you type the question and press send.

Try Qwen3 VL 235B A22B Instruct on your own question.

It opens in the chat with this model already selected.

Sign in to Upend

Use an account you already have, or get a one-time code by email.

Use a password instead

For accounts that already have a password. Forgotten it? An emailed code signs you in just the same.