Unlock exclusive introductory pricing for the newly launched Gemini 3.5 Flash.
Q

qwen3-vl-235b-a22b

GiriลŸ:$0.24/M
ร‡ฤฑktฤฑ:$0.96/M
BaฤŸlam:2M
Maksimum ร‡ฤฑktฤฑ:30K
Yayฤฑnlandฤฑ:Oct 1, 2025

qwen3-vl-235b-a22b is a multimodal model that unifies strong text generation with visual understanding for images and videos. Its Instruct variant optimizes instruction-following for general multimodal tasks. It excels in perception of real-world/synthetic categories, 2D/3D spatial grounding, and long-form visual comprehension, achieving competitive multimodal benchmark results.

Yeni
Ticari kullanฤฑm

qwen3-vl-235b-a22b iรงin Playground

qwen3-vl-235b-a22b'ฤฑn Playground'unu keลŸfedin โ€” modelleri test etmek ve sorgularฤฑ gerรงek zamanlฤฑ olarak รงalฤฑลŸtฤฑrmak iรงin etkileลŸimli bir ortam. Prompt'larฤฑ deneyin, parametreleri ayarlayฤฑn ve geliลŸtirmeyi hฤฑzlandฤฑrmak ve kullanฤฑm senaryolarฤฑnฤฑ doฤŸrulamak iรงin anฤฑnda yineleyin.

What is Qwen3-VL-235B-A22B

Qwen3-VL-235B-A22B is a high-capacity multimodal LLM from the Qwen (Alibaba) family. It combines a large MoE transformer backbone with cross-modal vision encoders and new positional/time encoding techniques to handle multi-image and long-duration video inputs, and to perform tasks such as visual question answering (VQA), long-document OCR, spatial/3D grounding, multimodal code generation, and agentic GUI control. The release includes both Instruct (task/few-shot tuned for instruction following) and Thinking (additional reasoning support and internal โ€œthinkโ€ mode) variants.

Main features (what makes Qwen3-VL-235B-A22B distinctive)

  • Large MoE design with high active capacity: a MoE stack that activates a subset of experts per request (โ‰ˆ22B active) to give more compute when needed while controlling inference cost.
  • Very long native context (256K) and scalable to ~1M: intended for book-length documents, hours of video, and multi-document workflows without aggressive chunking.
  • Advanced visual reasoning (spatial & temporal): Interleaved-MRoPE and DeepStack modules for timestamp alignment and fine-grained imageโ€“text fusion enabling video timeline queries and 3D grounding.
  • Improved OCR & document parsing: expanded OCR language support (advertised ~32 languages), stronger robustness to blur/tilt/low light and long, multi-page document structure parsing.
  • Visual agent + GUI automation: explicit agent capabilities to identify GUI elements, invoke functions or tools, and perform automation tasks on PC/mobile UIs.
  • Visual coding & multimodal program synthesis: can translate images/video/UI sketches into Draw.io/HTML/CSS/JS and assist in UI debugging.

How Qwen3-VL-235B-A22B compares to other models

Below are high-level comparisons to contemporaries; numbers and caps are taken from public provider/model pages and aggregator writeups.

  • Google Gemini 3 Pro โ€” Gemini emphasizes very large multimodal reasoning and agentic tool use; Google advertises 1M token context modes and deep product integrations. Gemini is positioned as a general leader in agentic multimodality (closed-source / proprietary), and often outperforms publicly available open models on some productized benchmarks. Qwen3-VL competes more directly as a high-capacity open-weight alternative optimized for OCR, video timeline alignment, and MoE cost tradeoffs.
  • Grok-4 Heavy (xAI) โ€” Grok-4 is another long-context, high-reasoning model family; some Grok variants list ~256K context windows and strong coding/math performance. Qwen3-VL and Grok-4 both target long-form reasoning; Qwen3-VL differentiates via heavy visual/video/OCR tooling and MoE scaling.
  • DeepSeek-R1 / DeepSeek family โ€” DeepSeek R1 emphasizes efficient training and competitive reasoning performance at lower inference cost; it is often used as an open alternative for reasoning/code tasks. Qwen3-VL targets stronger multimodal and spatial/video capabilities than R1โ€™s primary focus on text reasoning.

Representative use cases

  • Document parsing and large-scale OCR โ€” long, multi-page invoices, books, historical documents with multilingual text.
  • Video understanding & timeline queries โ€” summarize hours of recorded video, locate events by time, align text to video timestamps.
  • Visual question answering & multimodal assistants โ€” multi-turn image + text dialogs (customer support with screenshots, medical imaging notes).
  • GUI automation / visual agents โ€” detect UI elements and drive PC/mobile flows (automation, testing, assistive agents).
  • Multimodal code generation & UI prototyping โ€” convert mockups / images into HTML/CSS/JS or Draw.io diagrams.
  • Research & large-document analysis โ€” book-level summarization, multi-document synthesis with a single context.

How to access Qwen3 VL-235B-A22B API

Step 1: Sign Up for API Key

Log in to cometapi.com. If you are not our user yet, please register first. Sign into your CometAPI console. Get the access credential API key of the interface. Click โ€œAdd Tokenโ€ at the API token in the personal center, get the token key: sk-xxxxx and submit.

Step 2: Send Requests to Qwen3 VL-235B-A22B API

Select the โ€œQwen3-VL-235B-A22Bโ€ endpoint to send the API request and set the request body. The request method and request body are obtained from our website API doc. Our website also provides Apifox test for your convenience. Replace <YOUR_API_KEY> with your actual CometAPI key from your account. base url is Chat

Insert your question or request into the content fieldโ€”this is what the model will respond to . Process the API response to get the generated answer.

Step 3: Retrieve and Verify Results

Process the API response to get the generated answer. After processing, the API responds with the task status and output data.

qwen3-vl-235b-a22b iรงin Fiyatlandฤฑrma

qwen3-vl-235b-a22b iรงin รงeลŸitli bรผtรงelere ve kullanฤฑm ihtiyaรงlarฤฑna uygun rekabetรงi fiyatlandฤฑrmayฤฑ keลŸfedin. Esnek planlarฤฑmฤฑz sadece kullandฤฑฤŸฤฑnฤฑz kadar รถdeme yapmanฤฑzฤฑ saฤŸlar ve ihtiyaรงlarฤฑnฤฑz bรผyรผdรผkรงe kolayca รถlรงeklendirme imkanฤฑ sunar. qwen3-vl-235b-a22b'in maliyetleri yรถnetilebilir tutarken projelerinizi nasฤฑl geliลŸtirebileceฤŸini keลŸfedin.

ModelComet Price (USD / M Tokens)Official Price (USD / M Tokens)Discount
qwen3-vl-235b-a22b
GiriลŸ:$0.24/M
ร‡ฤฑktฤฑ:$0.96/M
GiriลŸ:$0.3/M
ร‡ฤฑktฤฑ:$1.2/M
-20%

qwen3-vl-235b-a22b iรงin รถrnek kod ve API

qwen3-vl-235b-a22b iรงin kapsamlฤฑ รถrnek kodlara ve API kaynaklarฤฑna eriลŸerek entegrasyon sรผrecinizi kolaylaลŸtฤฑrฤฑn. Ayrฤฑntฤฑlฤฑ dokรผmantasyonumuz adฤฑm adฤฑm rehberlik saฤŸlayarak projelerinizde qwen3-vl-235b-a22b'in tรผm potansiyelinden yararlanmanฤฑza yardฤฑmcฤฑ olur.

curl https://api.cometapi.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COMETAPI_KEY" \
  -d '{
    "model": "qwen3-vl-235b-a22b",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "Hello!"
      }
    ]
  }'

cURL Code Example

curl https://api.cometapi.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COMETAPI_KEY" \
  -d '{
    "model": "qwen3-vl-235b-a22b",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "Hello!"
      }
    ]
  }'

Python Code Example

from openai import OpenAI
import os

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com/v1"

client = OpenAI(base_url=BASE_URL, api_key=COMETAPI_KEY)

completion = client.chat.completions.create(
    model="qwen3-vl-235b-a22b",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Hello!"},
    ],
)

print(completion.choices[0].message.content)

JavaScript Code Example

import OpenAI from "openai";

// Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
const api_key = process.env.COMETAPI_KEY || "<YOUR_COMETAPI_KEY>";
const base_url = "https://api.cometapi.com/v1";

const openai = new OpenAI({
  apiKey: api_key,
  baseURL: base_url,
});

const completion = await openai.chat.completions.create({
  model: "qwen3-vl-235b-a22b",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "Hello!" },
  ],
});

console.log(completion.choices[0].message.content);

qwen3-vl-235b-a22b Sรผrรผmleri

qwen3-vl-235b-a22b'nฤฑn birden fazla anlฤฑk gรถrรผntรผye sahip olmasฤฑnฤฑn nedeni; gรผncellemeler sonrasฤฑ รงฤฑktฤฑ varyasyonlarฤฑ nedeniyle tutarlฤฑlฤฑk iรงin eski anlฤฑk gรถrรผntรผlere ihtiyaรง duyulmasฤฑ, geliลŸtiricilere uyum ve geรงiลŸ iรงin bir geรงiลŸ dรถnemi saฤŸlanmasฤฑ ve kullanฤฑcฤฑ deneyimini optimize etmek iรงin kรผresel veya bรถlgesel uรง noktalara karลŸฤฑlฤฑk gelen farklฤฑ anlฤฑk gรถrรผntรผler iรงerebilir. Sรผrรผmler arasฤฑndaki detaylฤฑ farklar iรงin lรผtfen resmi belgelere baลŸvurun.

Model namedescription
qwen3-vl-235b-a22bstandard
qwen3-vl-235b-a22b-thinkingthinking version