Qwen3 VL 235B A22B Instruct

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.

alibaba/qwen3-vl-235b-a22b-instruct

Organization: Alibaba
Family: qwen
Providers: 12
Context: 131,072
Output limit: 32,768
Knowledge: 2025-03-31
Release: 2025-09-23
Updated: 2025-09-23
Weights: Open
Input: text, image
Output: text
Capabilities: Tools, Structured, Temperature

Providers

Compare every cataloged offer for this model, including its provider, route, token limits, price, and capabilities.

18 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
Amazon Bedrock	qwen.qwen3-vl-235b-a22b	262,144	262,000	$0.53 / $2.66	No	Yes	Yes	View
Deep Infra	Qwen/Qwen3-VL-235B-A22B-Instruct	262,144	32,768	$0.20 / $0.88	No	Yes	Yes	View
Eden AI	qwen/qwen3-vl-235b-a22b-instruct	131,072	32,768	$0.40 / $1.60	No	Yes	Yes	View
Hugging Face	Qwen/Qwen3-VL-235B-A22B-Instruct	131,072	32,768	$0.30 / $1.50	No	Yes	Yes	View
Jalapeno Cloud	Qwen3-VL-235B-A22B-Instruct	129,024	32,768	$0.30 / $1.50	No	Yes	Yes	View
Kilo Gateway	qwen/qwen3-vl-235b-a22b-instruct	131,072	32,768	$0.26 / $1.04	No	Yes	Yes	View
LLM Gateway	deepinfra/qwen3-vl-235b-a22b-instruct	262,144	32,768	$0.20 / $0.88	No	Yes	No	View
LLM Gateway	novita/qwen3-vl-235b-a22b-instruct	131,072	32,768	$0.30 / $1.50	No	Yes	No	View
Merge Gateway	qwen/qwen3-vl-235b-a22b-instruct	131,072	32,768	$0.287 / $1.147	No	Yes	Yes	View
NanoGPT	qwen/qwen3-vl-235b-a22b-instruct	131,072	32,768	$0.30 / $1.90	No	No	No	View
Alibabavia OpenRouter	qwen/qwen3-vl-235b-a22b-instruct unknown	131,072	32,768	$0.26 / $1.04	No	Yes	Yes	View
DeepInfravia OpenRouter	qwen/qwen3-vl-235b-a22b-instruct fp8	262,144	16,384	$0.20 / $0.88	No	Yes	Yes	View
Novitavia OpenRouter	qwen/qwen3-vl-235b-a22b-instruct bf16	131,072	32,768	$0.30 / $1.50	No	Yes	Yes	View
Parasailvia OpenRouter	qwen/qwen3-vl-235b-a22b-instruct fp8	131,072	32,768	$0.21 / $1.90	No	Yes	Yes	View
PrimeIntellectvia OpenRouter	qwen/qwen3-vl-235b-a22b-instruct unknown	262,144	32,768	$0.30 / $1.50	No	Yes	Yes	View
Venicevia OpenRouter	qwen/qwen3-vl-235b-a22b-instruct fp8	128,000	16,384	$0.21 / $1.90	No	Yes	Yes	View
OrcaRouter	qwen/qwen3-vl-235b-a22b-instruct	131,072	32,768	$0.40 / $1.60	No	Yes	Yes	View
Vercel AI Gateway	alibaba/qwen3-vl-235b-a22b-instruct	131,072	129,024	$0.40 / $1.60	No	No	Yes	View

Prices and limits apply to each offer. A dash means the value is not provided. OpenRouter offers show the hosting provider and routing channel separately.

Pricing

Compare token rates and cache charges across available routes and hosting configurations.

USD per million tokens

Input from$0.20Per 1M input tokens

Output from$0.88Per 1M output tokens

Available offers18Across 12 providers

Lowest listed input and output prices across the offers above; they may belong to different providers. These are base rates, before conditional pricing, discounts or additional fees.

Endpoint price comparisonLatest observed snapshot · 6 lowest-priced configurations

InputOutput

DeepInfradeepinfra/fp8

$0.20$0.88

Parasailparasail/fp8

$0.21$1.90

Venicevenice/fp8

$0.21$1.90

Alibabaalibaba

$0.26$1.04

Novitanovita/bf16

$0.30$1.50

PrimeIntellectprimeintellect

$0.30$1.50

qwen/qwen3-vl-235b-a22b-instructObserved Oct 4, 2026, 11:31 AM UTC

Input: $0.21
Output: $1.90
Cache read: $0.10

Hosting prices through OpenRouter All endpoints and cache rates

Endpoint	Input / 1M	Output / 1M	Cache read / 1M	Cache write / 1M
Alibabaalibaba	$0.26	$1.04	—	—
DeepInfradeepinfra/fp8	$0.20	$0.88	$0.11	—
Novitanovita/bf16	$0.30	$1.50	—	—
Parasailparasail/fp8	$0.21	$1.90	$0.10	—
PrimeIntellectprimeintellect	$0.30	$1.50	—	—
Venicevenice/fp8	$0.21	$1.90	$0.10	—

Each row is a hosting configuration. Open an offer in Providers for its full pricing conditions and additional fees. Endpoint observations may differ from the routing catalog's base price.

Performance

Review reported latency and output-throughput percentiles from the latest observed rolling window.

Latest 30-minute observation

Time to first token · milliseconds

Latency distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Alibabaalibaba

623–2,290.5 ms

DeepInfradeepinfra/fp8

1,647–20,241 ms

Novitanovita/bf16

3,154–8,264.88 ms

Parasailparasail/fp8

956.5–7,761.03 ms

PrimeIntellectprimeintellect

1,538.5–36,546.6 ms

Venicevenice/fp8

2,029–8,618.92 ms

All latency percentiles 6 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba	623 ms	875.25 ms	1,077 ms	2,290.5 ms
DeepInfradeepinfra/fp8	1,647 ms	3,466 ms	6,412.2 ms	20,241 ms
Novitanovita/bf16	3,154 ms	3,605.75 ms	4,066 ms	8,264.88 ms
Parasailparasail/fp8	956.5 ms	1,421 ms	3,332 ms	7,761.03 ms
PrimeIntellectprimeintellect	1,538.5 ms	2,194.5 ms	3,802.9 ms	36,546.6 ms
Venicevenice/fp8	2,029 ms	2,243 ms	2,554.1 ms	8,618.92 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Alibabaalibaba

47–57 t/s

DeepInfradeepinfra/fp8

16–28.78 t/s

Novitanovita/bf16

31–50 t/s

Parasailparasail/fp8

41–79 t/s

PrimeIntellectprimeintellect

34–43.93 t/s

Venicevenice/fp8

22–82 t/s

All throughput percentiles 6 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba	47 t/s	51 t/s	54 t/s	57 t/s
DeepInfradeepinfra/fp8	16 t/s	19 t/s	22 t/s	28.78 t/s
Novitanovita/bf16	31 t/s	37 t/s	43 t/s	50 t/s
Parasailparasail/fp8	41 t/s	51 t/s	62 t/s	79 t/s
PrimeIntellectprimeintellect	34 t/s	43 t/s	43.3 t/s	43.93 t/s
Venicevenice/fp8	22 t/s	27 t/s	43 t/s	82 t/s

Percentiles describe the observed request distribution. These are snapshots of a rolling window, not a historical time series.

Uptime

Compare successful request rates across the latest 5-minute, 30-minute, and 24-hour observations.

Successful requests by window

Availability by endpointCurrent overlapping windows · 6 endpoints

Endpoint5 minutes30 minutes24 hours

AlibabaalibabaObserved Oct 4, 2026, 11:31 AM UTC

98.47%

98.99%

99.25%

DeepInfradeepinfra/fp8Observed Oct 4, 2026, 11:31 AM UTC

88.24%

93.9%

97.03%

Novitanovita/bf16Observed Oct 4, 2026, 11:31 AM UTC

100%

99.6%

96.85%

Parasailparasail/fp8Observed Oct 4, 2026, 11:31 AM UTC

98.21%

98.77%

99.22%

PrimeIntellectprimeintellectObserved Oct 4, 2026, 11:31 AM UTC

—

98.18%

Venicevenice/fp8Observed Oct 4, 2026, 11:31 AM UTC

99.57%

99.54%

97.8%

Reported by OpenRouter at the time of collection; rate-limited requests are excluded. A dash means no measurement was reported. This is not a live availability check.

Benchmarks

Review attributed evaluation results on each source's original scale.

External evaluations

Sources: Artificial Analysis and Design Arena, via OpenRouter. Scores retain their original scales and are not OpenAI Suite ratings.

Detailed evaluations

Evaluation scoresAccuracy or item F1 · higher is better

gpqa diamondQwen: Qwen3 VL 235B A22B Instruct

69.487%

tau bench verified airlineQwen: Qwen3 VL 235B A22B Instruct

40.889%

Evaluation	Configuration	Source	Results	Details
gpqa diamond	Qwen: Qwen3 VL 235B A22B Instructqwen/qwen3-vl-235b-a22b-instruct	OpenRouter	69.487%	Details
tau bench verified airline	Qwen: Qwen3 VL 235B A22B Instructqwen/qwen3-vl-235b-a22b-instruct	OpenRouter	40.889%	Details

Sources: Artificial Analysis, Design Arena and OpenRouter. Evaluation configurations and scales differ; results are not interchangeable. Updated Oct 4, 2026.

Technical details & API

Explore specifications, supported parameters, reasoning controls, and API identifiers for each OpenRouter route.

Current route specifications

qwen/qwen3-vl-235b-a22b-instructObserved Oct 4, 2026, 11:31 AM UTC

Context: 262,144
Maximum output: 32,768
Tokenizer: Qwen3
Input: text, image
Output: text
Moderation: No
Knowledge cutoff: 2025-03-31
Added to OpenRouter: 2025-09-23

Base URLhttps://openrouter.ai/api/v1

Model IDqwen/qwen3-vl-235b-a22b-instruct

Version IDqwen/qwen3-vl-235b-a22b-instruct

Weights IDQwen/Qwen3-VL-235B-A22B-Instruct

Supported parameters

frequency_penalty
logit_bias
logprobs
max_tokens
min_p
presence_penalty
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
top_k
top_logprobs
top_p

Default parameters

top p: 0.8
temperature: 0.7

Frequently asked questions

Answers based on the model metadata and serving offers currently in the catalog.

How much does Qwen3 VL 235B A22B Instruct cost?

Listed rates start at $0.20 per million input tokens and $0.88 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer Qwen3 VL 235B A22B Instruct?

There are 18 cataloged offers across 12 providers. Compare model IDs, prices and limits in Providers.

What is the context window?

The cataloged model context is 131,072 tokens. Each serving endpoint may apply a different limit.

Does it support tools and structured output?

Tool calling: Yes. Structured output: Yes. Support can vary by endpoint.

Resources

Hugging Face ↗

Catalog history · 2 recorded revisions

Metadata observations since this model was first cataloged. These are not model release versions.

Sep 14, 2026 · f156c3186a64
Sep 12, 2026 · 733cd94cf93e