Can I run Qwen3 VL 235B locally?

The weights are open and hosted on Hugging Face, so yes in principle, but 235B parameters with a MoE architecture requires substantial multi-GPU hardware; this is not a model you will run on a consumer workstation.

What does it cost and how long is the context?

Via API it is priced at $0.21 per million input tokens and $1.90 per million output tokens, with a 262,144-token context window; Qwen notes the architecture supports expansion to 1M tokens.

How does it differ from the Qwen3 VL Thinking model?

This is the Instruct edition, tuned for direct instruction-following across vision and text tasks; Qwen also offers a Thinking edition with an explicit reasoning-chain mode for harder STEM and logic problems.

Qwen3 VL 235B A22B Instruct

3377°

#llm #qwen #vision #agents #long context #multilingual

Run this model

Offers / Ads

OpenRouter Terminal / API

Pay per token

Chat with this model in the browser or call it from your code with one API key.

Qwen | 262,144 token context | $0.21/M input, $1.90/M output. Qwen3-VL-235B-A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video. The Instruct model targets general vision-language use (VQA, document parsing, chart/table extraction, multilingual OCR). The series emphasizes robust perception (recognition of diverse real-world and synthetic categories), spatial understanding (2D/3D grounding), and long-form visual comprehension, with competitive results on public multimodal benchmarks for both perception and…Qwen | 262,144 token context | $0.21/M input, $1.90/M output. Qwen3-VL-235B-A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video. The Instruct model targets general vision-language use (VQA, document parsing, chart/table extraction, multilingual OCR). The series emphasizes robust perception (recognition of diverse real-world and synthetic categories), spatial understanding (2D/3D grounding), and long-form visual comprehension, with competitive results on public multimodal benchmarks for both perception and reasoning. Beyond analysis, Qwen3-VL supports agentic interaction and tool use: it can follow complex instructions over multi-image, multi-turn dialogues; align text to video timelines for precise temporal queries; and operate GUI elements for automation tasks. The models also enable visual coding workflows—turning sketches or mockups into code and assisting with UI debugging—while maintaining strong text-only performance comparable to the flagship Qwen3 language models. This makes Qwen3-VL suitable for production scenarios spanning document AI, multilingual OCR, software/UI assistance, spatial/embodied tasks, and research on vision-language agents.

Guides for this model

Open weights vs open source vs API-only: what you are actually allowed to do

How much VRAM do I actually need for local AI?

What is a GGUF file? Quantized models explained (Q4, Q8, fp8, NF4)

All guides

Captain's notes

Written by Captain, the house editor

Notes from 2026-10-13

Qwen3 VL 235B A22B Instruct is Qwen's flagship open-weight vision-language model, a 235B MoE architecture that activates 22B parameters per forward pass. It handles images, video, documents, and GUI interaction in a single model, with OCR across 32 languages and spatial grounding that extends to 3D. At $0.21 per million input tokens it is a serious option for production document AI, visual agents, and multimodal research.

Known for

GUI agent: operates PC and mobile interfaces, invokes tools, completes multi-step tasks
Visual coding: generates HTML, CSS, JS and Draw.io diagrams directly from images or video
OCR across 32 languages, including degraded scans, rare scripts and long-document structure
Video temporal grounding: aligns text queries to precise timestamps in long video
3D spatial reasoning and 2D grounding for embodied and robotics-adjacent tasks

Watch out for

235B total parameters: self-hosting demands serious multi-GPU infrastructure; not a laptop model.
Output at $1.90 per million tokens is not cheap if you are generating long responses at volume.
Thinking edition exists separately; this Instruct variant does not include the reasoning-chain mode.

Best for

Prompts to try

Parse a messy invoice

Attach a photo or scan of an invoice. Extract the vendor name, invoice number, date, line items with quantities and unit prices, subtotal, tax, and total. Return the result as a JSON object. If any field is illegible or absent, set its value to null.

Sketch to working HTML

Attach a hand-drawn or wireframe sketch of a web page layout. Generate clean, self-contained HTML and CSS that reproduces the layout as closely as possible. Use semantic tags, a mobile-friendly viewport meta tag, and inline the CSS in a style block.

Locate objects in a scene

Attach an image of a room or outdoor scene. List every distinct object you can identify, give each one a bounding box as [x_min, y_min, x_max, y_max] in pixel coordinates relative to the image dimensions, and note any occlusions or depth relationships you can infer.

Compare with

Qwen3.7 Flash
API-only model
1,000,000 token context · $0.03/M input, $0.13/M output
Qwen3.8 Max (0902)
API-only model
1,000,000 token context · $2.00/M input, $6.00/M output
Qwen3.8 Omni Flash
API-only model
1,000,000 token context · $0.15/M input, $0.47/M output
Qwen3.5 Plus 2026-02-15
API-only model
1,000,000 token context · $0.26/M input, $1.56/M output

More text models from qwen

Questions people ask

Can I run Qwen3 VL 235B locally?: The weights are open and hosted on Hugging Face, so yes in principle, but 235B parameters with a MoE architecture requires substantial multi-GPU hardware; this is not a model you will run on a consumer workstation.
What does it cost and how long is the context?: Via API it is priced at $0.21 per million input tokens and $1.90 per million output tokens, with a 262,144-token context window; Qwen notes the architecture supports expansion to 1M tokens.
How does it differ from the Qwen3 VL Thinking model?: This is the Instruct edition, tuned for direct instruction-following across vision and text tasks; Qwen also offers a Thinking edition with an explicit reasoning-chain mode for harder STEM and logic problems.

Captain's verdict
The most capable open-weight vision-language model Qwen has shipped, and the one to reach for when documents, GUIs, video and spatial tasks all land on the same desk.

Model card on Hugging Face

Captain writes these notes from the model's own description and model card. Prices and context sizes are the provider's at the time of writing and can change.

Ranking badge

Updated Oct 22, 2026

Made this model? Show this badge on your page.

Paste it into your Hugging Face card or Civitai description.

Change colors / personalize / get code

Not this creator?

Get your own badge

How this score adds up

This model has no popularity data from any source yet, so it carries a stand-in number below 7,000° that keeps the lists complete. It never ranks above a model with data. The breakdown appears here as soon as a source reports it.

Glossary: the terms used here

Heat score: One number for how much a model is used and wanted right now, from 0° to 9,999°. Every list on this site that says hottest is sorted by it and nothing else.
Floor: The 7,000° every model with real data starts from. The scale runs from 0° to 9,999°; the part below the floor is reserved for stand-ins (models without data), so that a model with data always ranks above one without. The floor is the same for everyone and decides nothing; the degrees above it are the measurement.
Download-equivalents: Everything the score treats as a download, added up: downloads themselves, each torrent seeder as 1,000, each AI Horde image and each run on a generation site as one.
Log scale: Counting by tens, not by ones: going from 1,000 to 10,000 downloads adds the same degrees as going from 10,000 to 100,000. It keeps giants from crowding out everything else.
Likes: Thumbs up and likes from every source, worth half a download each on the log scale.
Adoption rate: How many downloads a day a model gets right now, from the difference between two readings of a source. A quarter of the weight of the download term.
Measured and estimated: Measured means two readings of a source exist and the difference is real. Estimated means only today's total is known and the past is spread over the model's life; the site says which one it is showing.
Seeder: Someone who keeps a copy of the weights online for others to download over a torrent, or a hosted mirror doing the same. Each one counts as 1,000 downloads.
AI Horde: A volunteer network that generates images for free on donated computers. The number of images it has made with a model is public, and each counts as one download.
Generation sites: Sites that host their own copies of a model and let people generate with it, such as PixAI, SeaArt and Yodayo. Their downloads count like Civitai's and each generation run as one download.
OpenRouter tokens: A text model served through an API, such as Claude or GPT, has no weights to download, so its usage is measured in tokens: the pieces of text it read and wrote through OpenRouter, a gateway many apps use, over the last 30 days. Every 10,000,000 tokens count as one download. Source: OpenRouter (openrouter.ai/rankings), CC BY 4.0.
Full model (checkpoint): The weights themselves, the file that generates on its own. It gets a bonus because add-ons depend on it.
LoRA and embedding: Small add-on files that steer someone else's weights toward a style, a character or a concept. They cannot run alone.
Family and weights: The base a model is built on, such as SD 1.5, SDXL, Flux or Wan. Every model inherits the release date of its family's weights.
Age: Degrees taken off for how long ago the family's weights came out: 150° a year, up to 750°. It keeps old downloads that had years to pile up from freezing the top of the list.
Family boost: A small, bounded nudge for LoRAs of a family this site is pushing forward right now, always listed in the breakdown.
Stand-in score: A stable number below the floor given to a model no source reports on yet, so lists stay complete. It never ranks above a model with data.
New! (grace period): A model less than 7 days old shows New! instead of a score and holds no place in the rankings. Its score is computed from day one, but a model one day old grows faster than anything, so without the wait the week's #1 would always be whatever came out yesterday rather than what held up for a week.
Trend: How a model's growth over a period compares with the typical model's growth over the same period. Faster than typical reads as a rise, slower as a fall.
Typical model: The middle of the catalog: half the models with a figure grew faster over the period, half slower.

Similar Models

Qwen3 VL 235B A22B Thinking

TEXT / LLM

o3 Pro

TEXT / LLM

AI Models for Image, Video & Text Generation

Qwen3 VL 235B A22B Instruct is a text model for the Text / LLM family listed on ModelTrends.ai, a read-only catalog of open source AI image, video and text models. Compare it with other Text / LLM models and check its heat score to see how it is trending.

More Text / LLM models llm qwen vision agents long context multilingual data extraction

Browse by Model Family