Qwen3.6 35B-A3B

Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per token. It uses a hybrid sparse mixture-of-experts architecture combining Gated DeltaNet linear attention with standard gated attention layers, enabling efficient inference at a fraction of the compute cost. The model supports a 262K token native context window (extensible to 1M via YaRN) and accepts text, image, and video inputs. It includes integrated thinking mode with reasoning traces preserved across multi-turn conversations, function calling, and structured output. Released under the Apache 2.0 license.

alibaba/qwen3.6-35b-a3b

Organization: Alibaba
Family: qwen
Providers: 38
Context: 262,144
Output limit: 65,536
Knowledge: —
Release: 2026-04-17
Updated: —
Weights: Open
Input: text, image/bmp, image/jpeg, image/png, image/tiff, image/webp, image/heic, video/mp4, video/x-msvideo, video/x-matroska, video/quicktime, video/x-flv, video/x-ms-wmv
Output: text
Capabilities: —

Providers

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

55 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
302.AI	qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
AIHubMix	qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
AKI.IO	qwen3.6-35b	256,000	32,768	— / —	—	—	—	View
Ambient	qwen/qwen3.6-35b-a3b	262,144	262,144	— / —	—	—	—	View
Boundless	qwen3.6	262,144	65,536	— / —	—	—	—	View
CoreWeave	Qwen/Qwen3.6-35B-A3B	262,144	262,144	— / —	—	—	—	View
Cortecs	qwen3.6-35b-a3b	262,000	32,768	— / —	—	—	—	View
Deep Infra	Qwen/Qwen3.6-35B-A3B	262,144	81,920	— / —	—	—	—	View
DevPass (LLM Gateway)	qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
EmpirioLabs AI	qwen3-6-35b-a3b	131,072	16,384	— / —	—	—	—	View
engy	qwen3.6-35b-a3b	208,192	8,192	— / —	—	—	—	View
evroc	Qwen/Qwen3.6-35B-A3B	262,144	65,536	— / —	—	—	—	View
GreenPT	qwen3.6-35b-a3b	262,144	32,768	— / —	—	—	—	View
Hetzner	Qwen/Qwen3.6-35B-A3B-FP8	262,144	262,144	— / —	—	—	—	View
Hugging Face	Qwen/Qwen3.6-35B-A3B	262,144	65,536	— / —	—	—	—	View
InferX	Qwen3.6-35B-A3B-FP8	262,000	65,536	— / —	—	—	—	View
InferX	Qwen3.6-35B-A3B-fp8-no-thinking	262,000	65,536	— / —	—	—	—	View
Kilo Gateway	qwen/qwen3.6-35b-a3b	262,144	235,929	— / —	—	—	—	View
LLM Gateway	alibaba/qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
LLM Gateway	novita/qwen3.6-35b-a3b	262,144	64,000	— / —	—	—	—	View
LLMTR	qwen3-6-35b	16,384	16,384	— / —	—	—	—	View
Merge Gateway	qwen/qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
NaN	qwen3.6	262,144	65,536	— / —	—	—	—	View
NanoGPT	qwen/qwen3.6-35b-a3b	1,048,576	16,384	— / —	—	—	—	View
NanoGPT	qwen/qwen3.6-35b-a3b:thinking	1,048,576	16,384	— / —	—	—	—	View
NanoGPT	TEE/qwen3.6-35b-a3b	262,144	262,144	— / —	—	—	—	View
NEAR AI Cloud	Qwen/Qwen3.6-35B-A3B-FP8	262,144	8,192	— / —	—	—	—	View
Neuralwatt	qwen3.6-35b-fast	131,056	131,056	— / —	—	—	—	View
Neuralwatt	qwen3.6-35b	131,056	131,056	— / —	—	—	—	View
Neuralwatt	qwen3.6-35b-flex	131,056	131,056	— / —	—	—	—	View
Novita AI	qwen/qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
AkashMLvia OpenRouter	qwen/qwen3.6-35b-a3b fp8	262,144	235,929	$0.10 / $0.90	Yes	Yes	Yes	View
AtlasCloudvia OpenRouter	qwen/qwen3.6-35b-a3b fp8	262,144	65,536	$0.186 / $1.11375	Yes	Yes	No	View
CoreWeavevia OpenRouter	qwen/qwen3.6-35b-a3b fp8	262,144	235,929	$0.25 / $1.25	Yes	No	Yes	View
Darkbloomvia OpenRouter	qwen/qwen3.6-35b-a3b fp4	262,144	32,768	$0.05 / $0.70	Yes	Yes	Yes	View
DeepInfravia OpenRouter	qwen/qwen3.6-35b-a3b fp8	262,144	65,536	$0.10 / $0.95	Yes	Yes	Yes	View
DekaLLMvia OpenRouter	qwen/qwen3.6-35b-a3b unknown	262,144	235,929	$0.10 / $1.00	Yes	Yes	Yes	View
Parasailvia OpenRouter	qwen/qwen3.6-35b-a3b fp8	262,144	235,929	$0.15 / $1.00	Yes	Yes	Yes	View
Phalavia OpenRouter	qwen/qwen3.6-35b-a3b unknown	262,144	235,929	$0.20 / $1.27	Yes	Yes	Yes	View
SiliconFlowvia OpenRouter	qwen/qwen3.6-35b-a3b fp8	262,144	235,929	$0.24 / $1.80	Yes	No	Yes	View
Venicevia OpenRouter	qwen/qwen3.6-35b-a3b fp8	256,000	65,536	$0.10 / $1.00	Yes	Yes	No	View
Opper	qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
OrcaRouter	qwen/qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
QVAC	qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	View
SaladCloud AI Gateway	qwen3.6-35b-a3b	262,144	262,144	— / —	—	—	—	View
Scaleway	qwen3.6-35b-a3b	128,000	16,384	— / —	—	—	—	View
SiliconFlow	Qwen/Qwen3.6-35B-A3B	262,144	262,144	— / —	—	—	—	View
SiliconFlow (China)	Qwen/Qwen3.6-35B-A3B	262,144	65,536	— / —	—	—	—	View
Thinking Machines	Qwen/Qwen3.6-35B-A3B	65,536	65,536	— / —	—	—	—	View
Thinking Machines	Qwen/Qwen3.6-35B-A3B:peft:262144	262,144	262,144	— / —	—	—	—	View
Umans AI	umans-flash	262,144	32,768	— / —	—	—	—	View
Umans AI Coding Plan	umans-flash	262,144	262,144	— / —	—	—	—	View
Umans AI Coding Plan	umans-qwen3.6-35b-a3b	262,144	262,144	— / —	—	—	—	View
Venice AI	qwen3-6-35b-a3b	256,000	65,536	— / —	—	—	—	View
Zenifra	alibaba/qwen3.6-35b-a3b	262,144	65,536	— / —	—	—	—	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$0.05Per 1M input tokens

Output from$0.70Per 1M output tokens

Available offers55Across 38 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 · 10 lowest-priced configurations

InputOutput

Darkbloomdarkbloom/fp4

$0.05$0.70

AkashMLakashml/fp8

$0.10$0.90

DeepInfradeepinfra/fp8

$0.10$0.95

DekaLLMdekallm

$0.10$1.00

Venicevenice/fp8

$0.10$1.00

Parasailparasail/fp8

$0.15$1.00

AtlasCloudatlas-cloud/fp8

$0.186$1.11375

Phalaphala

$0.20$1.27

SiliconFlowsiliconflow/fp8

$0.24$1.80

CoreWeavecoreweave/fp8

$0.25$1.25

qwen/qwen3.6-35b-a3bObserved Oct 10, 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
AkashMLakashml/fp8	$0.10	$0.90	$0.05	—
AtlasCloudatlas-cloud/fp8	$0.186	$1.11375	$0.186	—
CoreWeavecoreweave/fp8	$0.25	$1.25	$0.25	—
Darkbloomdarkbloom/fp4	$0.05	$0.70	$0.025	—
DeepInfradeepinfra/fp8	$0.10	$0.95	$0.10	—
DekaLLMdekallm	$0.10	$1.00	$0.05	—
Parasailparasail/fp8	$0.15	$1.00	$0.05	—
Phalaphala	$0.20	$1.27	$0.056	—
SiliconFlowsiliconflow/fp8	$0.24	$1.80	$0.15	—
Venicevenice/fp8	$0.10	$1.00	—	—

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

AkashMLakashml/fp8

467.5–1,953.63 ms

AtlasCloudatlas-cloud/fp8

3,531–14,735.32 ms

CoreWeavecoreweave/fp8

477–6,784.52 ms

Darkbloomdarkbloom/fp4

1,897.5–18,696.5 ms

DeepInfradeepinfra/fp8

21,555.5–68,046.99 ms

DekaLLMdekallm

802–9,611.72 ms

All latency percentiles 10 endpoints

Endpoint	P50	P75	P90	P99
AkashMLakashml/fp8	467.5 ms	600 ms	888.9 ms	1,953.63 ms
AtlasCloudatlas-cloud/fp8	3,531 ms	6,265 ms	8,926.8 ms	14,735.32 ms
CoreWeavecoreweave/fp8	477 ms	713 ms	2,770.8 ms	6,784.52 ms
Darkbloomdarkbloom/fp4	1,897.5 ms	3,725 ms	6,290 ms	18,696.5 ms
DeepInfradeepinfra/fp8	21,555.5 ms	34,996.5 ms	45,493.8 ms	68,046.99 ms
DekaLLMdekallm	802 ms	1,783.75 ms	3,955.2 ms	9,611.72 ms
Parasailparasail/fp8	658 ms	872 ms	1,360 ms	5,311.57 ms
Phalaphala	3,988.5 ms	9,532.75 ms	13,515.9 ms	24,329.62 ms
SiliconFlowsiliconflow/fp8	632 ms	1,654.75 ms	4,411.4 ms	12,123.14 ms
Venicevenice/fp8	651 ms	918 ms	1,631.8 ms	4,494.32 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

AkashMLakashml/fp8

68–128 t/s

AtlasCloudatlas-cloud/fp8

38–250.56 t/s

CoreWeavecoreweave/fp8

123–188.35 t/s

Darkbloomdarkbloom/fp4

80–185 t/s

DeepInfradeepinfra/fp8

5–16 t/s

DekaLLMdekallm

109–274.8 t/s

All throughput percentiles 10 endpoints

Endpoint	P50	P75	P90	P99
AkashMLakashml/fp8	68 t/s	80 t/s	94 t/s	128 t/s
AtlasCloudatlas-cloud/fp8	38 t/s	86 t/s	151 t/s	250.56 t/s
CoreWeavecoreweave/fp8	123 t/s	143 t/s	169.5 t/s	188.35 t/s
Darkbloomdarkbloom/fp4	80 t/s	124 t/s	149 t/s	185 t/s
DeepInfradeepinfra/fp8	5 t/s	8 t/s	11 t/s	16 t/s
DekaLLMdekallm	109 t/s	131 t/s	158 t/s	274.8 t/s
Parasailparasail/fp8	88 t/s	107 t/s	120 t/s	146.78 t/s
Phalaphala	35 t/s	44 t/s	49 t/s	68.6 t/s
SiliconFlowsiliconflow/fp8	96.5 t/s	159.75 t/s	183 t/s	228.97 t/s
Venicevenice/fp8	104 t/s	125 t/s	138 t/s	151.56 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 · 10 endpoints

Endpoint5 minutes30 minutes24 hours

AkashMLakashml/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

AtlasCloudatlas-cloud/fp8Observed Oct 10, 2026, 11:31 AM UTC

80%

53.33%

86.68%

CoreWeavecoreweave/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

99.9%

Darkbloomdarkbloom/fp4Observed Oct 10, 2026, 11:31 AM UTC

100%

DeepInfradeepinfra/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

76.67%

91.22%

DekaLLMdekallmObserved Oct 10, 2026, 11:31 AM UTC

100%

99.52%

Parasailparasail/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

99.86%

PhalaphalaObserved Oct 10, 2026, 11:31 AM UTC

—

80.95%

72.31%

SiliconFlowsiliconflow/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

94.44%

82.3%

Venicevenice/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

99.22%

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/qwen3.6-35b-a3bObserved Oct 10, 2026, 11:31 AM UTC

Benchmark profileArtificial Analysis · shared source scale from 0 to 50

Coding

41.9

Agentic

13.1

Intelligence

18.2

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: Qwen3.6 35B A3B

72.243%

gpqa diamondQwen: Qwen3.6 35B A3B

78.964%

Evaluation	Configuration	Source	Results	Details
Composite indices	Qwen3.6 35B A3B (Reasoning)qwen/qwen3.6-35b-a3b-20260415	Artificial Analysis	Intelligence 18.2 · Coding 41.9 · Agentic 13.1	Details
tau bench verified airline	Qwen: Qwen3.6 35B A3Bqwen/qwen3.6-35b-a3b-20260415	OpenRouter	72.243%	Details
gpqa diamond	Qwen: Qwen3.6 35B A3Bqwen/qwen3.6-35b-a3b-20260415	OpenRouter	78.964%	Details

Sources: Artificial Analysis, Design Arena and OpenRouter. Evaluation configurations and scales differ; results are not interchangeable. Updated Oct 24, 2026.

Apps & session costs

Compare observed median session costs for applications using this model, grouped by conversation length.

30-day sample · through Oct 18, 2026

Application	Model route	Turns	Median cost / session	Details
Hermes Agent	qwen/qwen3.6-35b-a3b-20260415	2–9 turns	$0.006755	Details
Hermes Agent	qwen/qwen3.6-35b-a3b-20260415	10–49 turns	$0.063761	Details

Source: OpenRouter, as of Oct 20, 2026. Published under CC BY 4.0. These are observed costs per session, not token prices or a forecast for your workload. Apps are compared separately.

Activity

Daily prompt and completion tokens reported for this model's routes in OpenRouter's top 50. Missing days have no published value.

Last 90 complete UTC days

Daily token usage1,016,781,597,456 reported tokens · 10 days with observations

386B0

2026-08-08
15,948,658,471 tokens

2026-08-09
26,659,873,315 tokens

2026-08-10
25,655,613,772 tokens

2026-08-13
35,160,314,706 tokens

2026-08-16
34,020,667,671 tokens

2026-10-11
57,549,182,336 tokens

2026-10-14
130,143,320,717 tokens

2026-10-15
167,353,760,522 tokens

2026-10-16
386,146,523,852 tokens

2026-10-17
138,143,682,094 tokens

Jul 26, 2026Oct 23, 2026

Daily observations · 10 days

Date (UTC)	Reported tokens	Reported routes
Oct 17, 2026	138,143,682,094	1
Oct 16, 2026	386,146,523,852	1
Oct 15, 2026	167,353,760,522	1
Oct 14, 2026	130,143,320,717	1
Oct 11, 2026	57,549,182,336	1
Aug 16, 2026	34,020,667,671	1
Aug 13, 2026	35,160,314,706	1
Aug 10, 2026	25,655,613,772	1
Aug 9, 2026	26,659,873,315	1
Aug 8, 2026	15,948,658,471	1

Source: OpenRouter (openrouter.ai/rankings), as of Oct 24, 2026. CC BY 4.0. Totals include only individually published routes. Tokenizers differ by provider; the aggregated “other” category is never assigned to a model.

Technical details & API

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

Current route specifications

qwen/qwen3.6-35b-a3bObserved Oct 10, 2026, 11:31 AM UTC

Context: 262,144
Maximum output: 235,929
Tokenizer: Qwen
Input: text, image, video
Output: text
Moderation: No
Added to OpenRouter: 2026-04-27

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

Model IDqwen/qwen3.6-35b-a3b

Version IDqwen/qwen3.6-35b-a3b-20260415

Weights IDQwen/Qwen3.6-35B-A3B

Reasoning

Required: No
Enabled by default: Yes
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 Qwen3.6 35B-A3B cost?

Listed rates start at $0.05 per million input tokens and $0.70 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer Qwen3.6 35B-A3B?

There are 55 cataloged offers across 38 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: Not reported. Structured output: Not reported. Support can vary by endpoint.

Resources

Hugging Face ↗

Catalog history · 3 recorded revisions

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

Oct 20, 2026 · f394c5225049
Sep 14, 2026 · 33a7a88c5273
Sep 12, 2026 · 5eeaa36e40c5