Qwen3.5 397B-A17B

The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. It delivers state-of-the-art performance comparable to leading-edge models across a wide range of tasks, including language understanding, logical reasoning, code generation, agent-based tasks, image understanding, video understanding, and graphical user interface (GUI) interactions. With its robust code-generation and agent capabilities, the model exhibits strong generalization across diverse agent.

alibaba/qwen3.5-397b-a17b

Organization: Alibaba
Family: qwen
Providers: 22
Context: 262,144
Output limit: 65,536
Knowledge: —
Release: 2026-02-15
Updated: 2026-02-15
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.

34 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
AIHubMix	qwen3.5-397b-a17b	262,144	65,536	$0.1644 / $0.9864	Yes	Yes	Yes	View
Chutes	Qwen/Qwen3.5-397B-A17B-TEE	262,144	65,536	$0.45 / $3.00	Yes	Yes	Yes	View
Cloudflare AI Gateway	alibaba/qwen3.5-397b-a17b	262,144	65,536	$0.60 / $3.60	Yes	Yes	Yes	View
Cortecs	qwen3.5-397b-a17b	262,000	262,000	$0.668 / $4.01	Yes	Yes	—	View
CrofAI	qwen3.5-397b-a17b	262,144	262,144	$0.35 / $1.75	Yes	Yes	Yes	View
DevPass (LLM Gateway)	qwen35-397b-a17b	262,144	65,536	$0.60 / $3.60	Yes	Yes	Yes	View
EmpirioLabs AI	qwen3-5-397b-a17b	256,000	64,000	$0.172 / $1.032	Yes	Yes	Yes	View
GreenPT	qwen3.5-397b-a17b	262,144	16,384	$0.798 / $4.959	Yes	Yes	Yes	View
Infomaniak	Qwen/Qwen3.5-397B-A17B-FP8	200,000	65,536	$0.99 / $4.46	Yes	Yes	Yes	View
Jalapeno Cloud	Qwen3.5-397B-A17B	262,144	65,536	$0.60 / $3.60	Yes	Yes	Yes	View
Kilo Gateway	qwen/qwen3.5-397b-a17b	262,144	235,929	$0.39 / $2.34	Yes	Yes	Yes	View
LLMTR	qwen/qwen3.5-397b-a17b	256,000	65,536	$0.60 / $3.60	Yes	Yes	Yes	View
Merge Gateway	qwen/qwen3.5-397b-a17b	131,072	32,768	$0.172 / $1.032	Yes	Yes	Yes	View
NanoGPT	TEE/qwen3.5-397b-a17b	262,144	65,536	$0.55 / $3.50	No	No	No	View
NanoGPT	qwen/qwen3.5-397b-a17b	258,048	65,536	$0.60 / $3.60	No	Yes	No	View
NanoGPT	qwen/qwen3.5-397b-a17b:thinking	258,048	65,536	$0.60 / $3.60	Yes	Yes	No	View
Ofox	bailian/qwen3.5-397b-a17b	256,000	64,000	$0.55 / $3.50	Yes	Yes	Yes	View
Ofox	qwen/qwen3.5-397b-a17b	256,000	64,000	$0.55 / $3.50	Yes	Yes	Yes	View
Alibabavia OpenRouter	qwen/qwen3.5-397b-a17b unknown	262,144	65,536	$0.39 / $2.34	Yes	Yes	No	View
AtlasCloudvia OpenRouter	qwen/qwen3.5-397b-a17b fp8	262,144	65,536	$0.55 / $3.50	Yes	Yes	Yes	View
DeepInfravia OpenRouter	qwen/qwen3.5-397b-a17b fp8	262,144	81,920	$0.45 / $3.00	Yes	Yes	Yes	View
DigitalOceanvia OpenRouter	qwen/qwen3.5-397b-a17b unknown	131,072	117,964	$0.55 / $3.50	Yes	Yes	Yes	View
GMICloudvia OpenRouter	qwen/qwen3.5-397b-a17b fp8	262,144	235,929	$0.60 / $3.60	Yes	Yes	No	View
Novitavia OpenRouter	qwen/qwen3.5-397b-a17b unknown	262,144	65,536	$0.60 / $3.60	Yes	Yes	No	View
Parasailvia OpenRouter	qwen/qwen3.5-397b-a17b fp8	262,144	235,929	$0.50 / $3.60	Yes	Yes	Yes	View
Phalavia OpenRouter	qwen/qwen3.5-397b-a17b unknown	262,144	235,929	$0.55 / $3.50	Yes	Yes	Yes	View
StreamLakevia OpenRouter	qwen/qwen3.5-397b-a17b unknown	256,000	64,000	$0.60 / $3.60	Yes	No	Yes	View
Venicevia OpenRouter	qwen/qwen3.5-397b-a17b unknown	128,000	32,768	$0.75 / $4.50	Yes	Yes	Yes	View
OrcaRouter	qwen/qwen3.5-397b-a17b	262,144	65,536	$0.172 / $1.032	Yes	Yes	Yes	View
Scaleway	qwen3.5-397b-a17b	256,000	16,384	$0.60 / $3.60	Yes	Yes	Yes	View
SiliconFlow	Qwen/Qwen3.5-397B-A17B	262,144	262,144	$0.39 / $2.34	No	Yes	Yes	View
TokenGo	qwen/qwen3.5-397b-a17b	262,144	65,536	$0.40 / $2.65	Yes	Yes	Yes	View
UnoRouter	qwen3.5-397b-a17b:free	262,144	65,536	$0.00 / $0.00	Yes	Yes	Yes	View
Venice AI	qwen3-5-397b-a17b	128,000	32,768	$0.75 / $4.50	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$0.00Per 1M input tokens

Output from$0.00Per 1M output tokens

Available offers34Across 22 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

Alibabaalibaba

$0.39$2.34

DeepInfradeepinfra/fp8

$0.45$3.00

Parasailparasail/fp8

$0.50$3.60

AtlasCloudatlas-cloud/fp8

$0.55$3.50

DigitalOceandigitalocean

$0.55$3.50

Phalaphala

$0.55$3.50

GMICloudgmicloud/fp8

$0.60$3.60

Novitanovita

$0.60$3.60

StreamLakestreamlake

$0.60$3.60

Venicevenice

$0.75$4.50

qwen/qwen3.5-397b-a17bObserved Oct 4, 2026, 11:31 AM UTC

Input: $0.55
Output: $3.50
Cache read: $0.225

Hosting prices through OpenRouter All endpoints and cache rates

Endpoint	Input / 1M	Output / 1M	Cache read / 1M	Cache write / 1M
Alibabaalibaba	$0.39	$2.34	—	—
AtlasCloudatlas-cloud/fp8	$0.55	$3.50	$0.55	—
DeepInfradeepinfra/fp8	$0.45	$3.00	$0.22	—
DigitalOceandigitalocean	$0.55	$3.50	$0.11	—
GMICloudgmicloud/fp8	$0.60	$3.60	—	—
Novitanovita	$0.60	$3.60	—	—
Parasailparasail/fp8	$0.50	$3.60	$0.30	—
Phalaphala	$0.55	$3.50	$0.225	—
StreamLakestreamlake	$0.60	$3.60	$0.12	—
Venicevenice	$0.75	$4.50	—	—

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

925.5–6,964.41 ms

AtlasCloudatlas-cloud/fp8

1,396.5–7,656 ms

DeepInfradeepinfra/fp8

1,355–11,795.6 ms

DigitalOceandigitalocean

1,197–14,383.84 ms

GMICloudgmicloud/fp8

2,105–5,200.44 ms

Novitanovita

1,310–4,707 ms

All latency percentiles 10 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba	925.5 ms	1,083.25 ms	1,384.9 ms	6,964.41 ms
AtlasCloudatlas-cloud/fp8	1,396.5 ms	1,933.75 ms	2,292.5 ms	7,656 ms
DeepInfradeepinfra/fp8	1,355 ms	2,779 ms	5,819 ms	11,795.6 ms
DigitalOceandigitalocean	1,197 ms	2,420 ms	6,709.2 ms	14,383.84 ms
GMICloudgmicloud/fp8	2,105 ms	2,632 ms	2,995.8 ms	5,200.44 ms
Novitanovita	1,310 ms	1,558.5 ms	1,959 ms	4,707 ms
Parasailparasail/fp8	1,056.5 ms	1,370.75 ms	1,742.2 ms	3,195.14 ms
Phalaphala	1,567 ms	2,540 ms	3,860.2 ms	43,646.96 ms
StreamLakestreamlake	549 ms	982 ms	1,657 ms	3,431.68 ms
Venicevenice	5,720 ms	9,798.75 ms	13,372.1 ms	16,301.74 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Alibabaalibaba

58–99.55 t/s

AtlasCloudatlas-cloud/fp8

49–87 t/s

DeepInfradeepinfra/fp8

44–83 t/s

DigitalOceandigitalocean

21–69.37 t/s

GMICloudgmicloud/fp8

88–96.65 t/s

Novitanovita

52–127.5 t/s

All throughput percentiles 10 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba	58 t/s	75 t/s	87.5 t/s	99.55 t/s
AtlasCloudatlas-cloud/fp8	49 t/s	64 t/s	74 t/s	87 t/s
DeepInfradeepinfra/fp8	44 t/s	56 t/s	67 t/s	83 t/s
DigitalOceandigitalocean	21 t/s	33 t/s	46 t/s	69.37 t/s
GMICloudgmicloud/fp8	88 t/s	91.25 t/s	93.5 t/s	96.65 t/s
Novitanovita	52 t/s	70 t/s	83 t/s	127.5 t/s
Parasailparasail/fp8	52.5 t/s	74 t/s	93.3 t/s	132.59 t/s
Phalaphala	45 t/s	59 t/s	66 t/s	74 t/s
StreamLakestreamlake	63 t/s	79 t/s	97 t/s	149 t/s
Venicevenice	21 t/s	36 t/s	49 t/s	79 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

AlibabaalibabaObserved Oct 4, 2026, 11:31 AM UTC

—

99.38%

97.47%

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

—

99.21%

94.76%

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

100%

99.45%

DigitalOceandigitaloceanObserved Oct 4, 2026, 11:31 AM UTC

98.16%

97.22%

95.23%

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

—

90.26%

NovitanovitaObserved Oct 4, 2026, 11:31 AM UTC

—

95.14%

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

100%

99.29%

PhalaphalaObserved Oct 4, 2026, 11:31 AM UTC

98.86%

99.43%

98.25%

StreamLakestreamlakeObserved Oct 4, 2026, 11:31 AM UTC

99.45%

99.84%

97.45%

VeniceveniceObserved Oct 4, 2026, 11:31 AM UTC

—

71.11%

84.66%

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/qwen3.5-397b-a17bObserved Oct 4, 2026, 11:31 AM UTC

Benchmark profileArtificial Analysis · shared source scale from 0 to 50

Coding

48.2

Agentic

8.3

Intelligence

18.4

Design Arena	Category	Elo	Win rate	Rank
models	3d	1,176	56.7%	55
models	codecategories	1,186	52.6%	58
models	dataviz	1,184	53.2%	60
models	gamedev	1,153	50.2%	67
models	svg	1,139	56.1%	49
models	uicomponent	1,170	51.4%	68
models	website	1,197	52.5%	59

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: Qwen3.5 397B A17B

84.238%

tau bench verified airlineQwen: Qwen3.5 397B A17B

76.917%

Evaluation	Configuration	Source	Results	Details
gpqa diamond	Qwen: Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	OpenRouter	84.238%	Details
Composite indices	Qwen3.5 397B A17B (Reasoning)qwen/qwen3.5-397b-a17b-20260216	Artificial Analysis	Intelligence 18.4 · Coding 48.2 · Agentic 8.3	Details
models · uicomponent	Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	Design Arena	1,170 Elo	Details
models · dataviz	Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	Design Arena	1,184 Elo	Details
models · codecategories	Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	Design Arena	1,186 Elo	Details
tau bench verified airline	Qwen: Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	OpenRouter	76.917%	Details
models · 3d	Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	Design Arena	1,176 Elo	Details
models · gamedev	Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	Design Arena	1,153 Elo	Details
models · svg	Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	Design Arena	1,139 Elo	Details
models · website	Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b-20260216	Design Arena	1,197 Elo	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/qwen3.5-397b-a17bObserved 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: 2026-02-16

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

Model IDqwen/qwen3.5-397b-a17b

Version IDqwen/qwen3.5-397b-a17b-20260216

Weights IDQwen/Qwen3.5-397B-A17B

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: 0.6

Frequently asked questions

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

How much does Qwen3.5 397B-A17B cost?

Listed rates start at $0.00 per million input tokens and $0.00 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer Qwen3.5 397B-A17B?

There are 34 cataloged offers across 22 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 · 0607fe244a54
Sep 12, 2026 · 545b2781378a