Qwen3.5 35B-A3B

The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency. Its overall performance is comparable to that of the Qwen3.5-27B.

alibaba/qwen3.5-35b-a3b

Organization: Alibaba
Family: qwen
Providers: 17
Context: 262,144
Output limit: 65,536
Knowledge: —
Release: 2026-02-23
Updated: 2026-02-23
Weights: Open
Input: text, image, video, audio
Output: text
Capabilities: Tools, Reasoning, Structured, Temperature

Providers

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

25 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
302.AI	qwen3.5-35b-a3b	262,144	65,536	$0.06 / $0.46	Yes	Yes	Yes	View
AIHubMix	qwen3.5-35b-a3b	262,144	65,536	$0.0564 / $0.4512	Yes	Yes	Yes	View
CoreWeave	Qwen/Qwen3.5-35B-A3B	262,144	262,144	$0.25 / $1.25	Yes	Yes	Yes	View
DevPass (LLM Gateway)	qwen3.5-35b-a3b	262,144	65,536	$0.25 / $2.00	Yes	Yes	Yes	View
EmpirioLabs AI	qwen3-5-35b-a3b	256,000	64,000	$0.057 / $0.459	Yes	Yes	Yes	View
Hugging Face	Qwen/Qwen3.5-35B-A3B	262,144	65,536	$0.25 / $2.00	Yes	Yes	Yes	View
Jalapeno Cloud	Qwen3.5-35B-A3B	262,144	65,536	$0.25 / $2.00	Yes	Yes	Yes	View
Kilo Gateway	qwen/qwen3.5-35b-a3b	262,144	235,929	$0.1625 / $1.30	Yes	Yes	Yes	View
LLM Gateway	alibaba/qwen3.5-35b-a3b	262,144	65,536	$0.25 / $2.00	Yes	Yes	No	View
Merge Gateway	qwen/qwen3.5-35b-a3b	131,072	32,768	$0.057 / $0.459	Yes	Yes	Yes	View
NanoGPT	qwen/qwen3.5-35b-a3b	260,096	65,536	$0.225 / $1.80	No	Yes	No	View
NanoGPT	qwen/qwen3.5-35b-a3b:thinking	260,096	65,536	$0.225 / $1.80	Yes	Yes	No	View
Ofox	bailian/qwen3.5-35b-a3b	262,144	65,536	$0.29 / $1.83	Yes	Yes	Yes	View
Ofox	qwen/qwen3.5-35b-a3b	256,000	64,000	$0.29 / $1.83	Yes	Yes	Yes	View
Alibabavia OpenRouter	qwen/qwen3.5-35b-a3b unknown	262,144	65,536	$0.1625 / $1.30	Yes	Yes	Yes	View
AtlasCloudvia OpenRouter	qwen/qwen3.5-35b-a3b fp8	262,144	65,536	$0.225 / $1.80	Yes	Yes	Yes	View
Darkbloomvia OpenRouter	qwen/qwen3.5-35b-a3b fp4	262,144	32,768	$0.08 / $0.75	Yes	Yes	Yes	View
DeepInfravia OpenRouter	qwen/qwen3.5-35b-a3b fp8	262,144	81,920	$0.14 / $1.00	Yes	No	Yes	View
Parasailvia OpenRouter	qwen/qwen3.5-35b-a3b fp8	262,144	235,929	$0.15 / $1.00	Yes	Yes	Yes	View
SiliconFlowvia OpenRouter	qwen/qwen3.5-35b-a3b fp8	262,144	235,929	$0.24 / $1.80	Yes	No	Yes	View
Venicevia OpenRouter	qwen/qwen3.5-35b-a3b unknown	256,000	16,384	$0.3125 / $1.25	Yes	Yes	Yes	View
OrcaRouter	qwen/qwen3.5-35b-a3b	262,144	65,536	$0.057 / $0.459	Yes	Yes	Yes	View
Requesty	qwen3.5-35b-a3b	262,144	262,144	$0.14 / $1.00	Yes	Yes	Yes	View
SiliconFlow	Qwen/Qwen3.5-35B-A3B	262,144	262,144	$0.24 / $1.80	No	Yes	Yes	View
Venice AI	qwen3-5-35b-a3b	256,000	16,384	$0.3125 / $1.25	Yes	Yes	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$0.0564Per 1M input tokens

Output from$0.4512Per 1M output tokens

Available offers25Across 17 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 · 7 lowest-priced configurations

InputOutput

Darkbloomdarkbloom/fp4

$0.08$0.75

DeepInfradeepinfra/fp8

$0.14$1.00

Parasailparasail/fp8

$0.15$1.00

Alibabaalibaba

$0.1625$1.30

AtlasCloudatlas-cloud/fp8

$0.225$1.80

SiliconFlowsiliconflow/fp8

$0.24$1.80

Venicevenice

$0.3125$1.25

qwen/qwen3.5-35b-a3bObserved Oct 4, 2026, 11:31 AM UTC

Input: $0.15
Output: $1.00
Cache read: $0.05

Hosting prices through OpenRouter All endpoints and cache rates

Endpoint	Input / 1M	Output / 1M	Cache read / 1M	Cache write / 1M
Alibabaalibaba	$0.1625	$1.30	—	—
AtlasCloudatlas-cloud/fp8	$0.225	$1.80	$0.225	—
Darkbloomdarkbloom/fp4	$0.08	$0.75	$0.04	—
DeepInfradeepinfra/fp8	$0.14	$1.00	$0.05	—
Parasailparasail/fp8	$0.15	$1.00	$0.05	—
SiliconFlowsiliconflow/fp8	$0.24	$1.80	$0.15	—
Venicevenice	$0.3125	$1.25	$0.15625	—

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

655–1,464.86 ms

AtlasCloudatlas-cloud/fp8

1,008–4,037.39 ms

Darkbloomdarkbloom/fp4

1,245–21,865 ms

DeepInfradeepinfra/fp8

666–2,151.6 ms

Parasailparasail/fp8

1,159–4,653.12 ms

SiliconFlowsiliconflow/fp8

1,740–4,722.86 ms

All latency percentiles 7 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba	655 ms	844 ms	977 ms	1,464.86 ms
AtlasCloudatlas-cloud/fp8	1,008 ms	1,180.75 ms	1,550.5 ms	4,037.39 ms
Darkbloomdarkbloom/fp4	1,245 ms	1,892 ms	3,604 ms	21,865 ms
DeepInfradeepinfra/fp8	666 ms	915.5 ms	1,150 ms	2,151.6 ms
Parasailparasail/fp8	1,159 ms	2,378 ms	3,078.8 ms	4,653.12 ms
SiliconFlowsiliconflow/fp8	1,740 ms	2,416.75 ms	3,874.8 ms	4,722.86 ms
Venicevenice	1,004.5 ms	1,509.5 ms	1,862.5 ms	2,727.7 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Alibabaalibaba

125–191.66 t/s

AtlasCloudatlas-cloud/fp8

64–86 t/s

Darkbloomdarkbloom/fp4

85–147.56 t/s

DeepInfradeepinfra/fp8

61–114.42 t/s

Parasailparasail/fp8

143–243.72 t/s

SiliconFlowsiliconflow/fp8

41–76.4 t/s

All throughput percentiles 7 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba	125 t/s	143.5 t/s	159.8 t/s	191.66 t/s
AtlasCloudatlas-cloud/fp8	64 t/s	69 t/s	75 t/s	86 t/s
Darkbloomdarkbloom/fp4	85 t/s	102 t/s	117 t/s	147.56 t/s
DeepInfradeepinfra/fp8	61 t/s	66 t/s	68 t/s	114.42 t/s
Parasailparasail/fp8	143 t/s	218.5 t/s	233.2 t/s	243.72 t/s
SiliconFlowsiliconflow/fp8	41 t/s	55.5 t/s	71 t/s	76.4 t/s
Venicevenice	150 t/s	168.25 t/s	186 t/s	217.09 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 · 7 endpoints

Endpoint5 minutes30 minutes24 hours

AlibabaalibabaObserved Oct 4, 2026, 11:31 AM UTC

—

68.33%

AtlasCloudatlas-cloud/fp8Observed Oct 4, 2026, 11:31 AM UTC

100%

98.96%

Darkbloomdarkbloom/fp4Observed Oct 4, 2026, 11:31 AM UTC

—

99.71%

99.84%

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

100%

99.63%

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

—

99.15%

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

—

95.13%

VeniceveniceObserved Oct 4, 2026, 11:31 AM UTC

100%

99.92%

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

qwen/qwen3.5-35b-a3bObserved Oct 4, 2026, 11:31 AM UTC

Benchmark profileArtificial Analysis · shared source scale from 0 to 40

Coding

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

tau bench verified airlineQwen: Qwen3.5-35B-A3B

65.038%

gpqa diamondQwen: Qwen3.5-35B-A3B

79.84%

Evaluation	Configuration	Source	Results	Details
tau bench verified airline	Qwen: Qwen3.5-35B-A3Bqwen/qwen3.5-35b-a3b-20260224	OpenRouter	65.038%	Details
Composite indices	Qwen3.5 35B A3B (Non-reasoning)qwen/qwen3.5-35b-a3b-20260224	Artificial Analysis	Intelligence — · Coding 37 · Agentic —	Details
gpqa diamond	Qwen: Qwen3.5-35B-A3Bqwen/qwen3.5-35b-a3b-20260224	OpenRouter	79.84%	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/qwen3.5-35b-a3bObserved Oct 4, 2026, 11:31 AM UTC

Context: 262,144
Maximum output: 235,929
Tokenizer: Qwen3
Input: text, image, video
Output: text
Moderation: No
Added to OpenRouter: 2026-02-25

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

Model IDqwen/qwen3.5-35b-a3b

Version IDqwen/qwen3.5-35b-a3b-20260224

Weights IDQwen/Qwen3.5-35B-A3B

Reasoning

Required: No
Enabled by default: Not reported
Default effort: —
Supported efforts: —

Supported parameters

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

Default parameters

top k: 20
top p: 0.95
temperature: 1

Frequently asked questions

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

How much does Qwen3.5 35B-A3B cost?

Listed rates start at $0.0564 per million input tokens and $0.4512 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer Qwen3.5 35B-A3B?

There are 25 cataloged offers across 17 providers. Compare model IDs, prices and limits in Providers.

What is the context window?

The cataloged model context is 262,144 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 · 742decf79134
Sep 12, 2026 · 6dcec55cbd75