Qwen: Qwen3.5 397B A17B

apache-2.0397B / 17B

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.

Context
262K
Max output
236K
Input price
$0.5/1M tokens
Output price
$3.6/1M tokens

Capabilities

Reasoning & Interaction
Reasoning
Tool Calling
Output Control
Structured Output
JSON Mode
Infrastructure
Streaming
Batch Processing
Fine-tuning

Details

Model IDqwen/qwen3.5-397b-a17b
Provider OpenRouter
Creator Qwen
Familyqwen3.5
Licenseapache-2.0
Status—
Typechat
Input modalitiestext, image, video
Output modalitiestext
Open weightYes
Parameters397B / 17B
Architecture—
Context window262K tokens
Max input—
Max output236K tokens
Reasoning tokens—
Knowledge cutoff—
Release date
Sourceofficial

Tools

Function Callingfunction_calling
Call external functions and APIs

Pricing

Input
$0.5
Output
$3.6
Cache write
—
Cache read
$0.3
Batch in
—
Batch out
—

Family Comparison: qwen3.5

ModelContextMax outPricing
Qwen: Qwen3.5-Flash1M66K$0.07$0.26
Qwen: Qwen3.5-35B-A3B262K82K$0.14$1
Qwen: Qwen3.5-9B (batch)262K236K$0.17$0.25
Qwen: Qwen3.5-9B262K236K$0.17$0.25
Qwen: Qwen3.5-122B-A10B262K66K$0.26$2.08
Qwen: Qwen3.5-27B262K82K$0.26$2.6
Qwen: Qwen3.5 Plus 2026-02-151M66K$0.26$1.56
Qwen: Qwen3.5 Plus 2026-04-201M66K$0.3$1.8
Qwen: Qwen3.5 397B A17B262K236K$0.5$3.6

API

GET/v1/models/openrouter/qwen/qwen3.5-397b-a17b

Qwen: Qwen3.5 397B A17B

qwen/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.

Changes · 18 entries
qwen/qwen3.5-397b-a17bupdateca7a400Nov 2, 2026, 03:02 AM
max_output_tokens65536→235929
pricinginput$0.39→$0.5output$2.34→$3.6cached input$0.195→$0.3
qwen/qwen3.5-397b-a17bupdatef054294Oct 22, 2026, 06:37 AM
max_output_tokens65536→235929
pricinginput$0.39→$0.5output$2.34→$3.6
qwen/qwen3.5-397b-a17bupdate42a9895Oct 20, 2026, 06:39 AM
max_output_tokens235929→65536
pricinginput$0.5→$0.39output$3.6→$2.34
qwen/qwen3.5-397b-a17bupdate0d5a9a1Oct 14, 2026, 06:38 AM
pricinginput$0.55→$0.5output$3.5→$3.6cached input$0.225→$0.3
qwen/qwen3.5-397b-a17bupdate70614e7Oct 9, 2026, 06:39 AM
max_output_tokens81920→235929
pricinginput$0.45→$0.55output$3→$3.5cached input$0.22→$0.225
qwen/qwen3.5-397b-a17bupdate00710a0Oct 6, 2026, 06:38 AM
max_output_tokens235929→81920
pricinginput$0.55→$0.45output$3.5→$3cached input$0.225→$0.22
qwen/qwen3.5-397b-a17bupdate79684d6Sep 9, 2026, 06:38 AM
max_output_tokens65536→235929
pricinginput$0.39→$0.55output$2.34→$3.5
qwen/qwen3.5-397b-a17bupdatea5ce2fdSep 7, 2026, 06:48 AM
max_output_tokens235929→65536
pricinginput$0.55→$0.39output$3.5→$2.34
qwen/qwen3.5-397b-a17bupdated79e481Sep 3, 2026, 06:38 AM
max_output_tokens65536→235929
pricinginput$0.39→$0.55output$2.34→$3.5cached input$0.045→$0.225
qwen/qwen3.5-397b-a17bupdatebe35184Aug 17, 2026, 04:02 AM
pricinginput$0.45→$0.39output$3→$2.34
qwen/qwen3.5-397b-a17bupdate3cddff5Aug 13, 2026, 02:54 AM
pricinginput$0.39→$0.45output$2.34→$3cached input$0.111→$0.045
qwen/qwen3.5-397b-a17bupdate20c73acJul 17, 2026, 02:34 AM
context_window256000→262144
pricinginput$0.385→$0.39output$2.45→$2.34
qwen/qwen3.5-397b-a17bupdate53ef452Jul 9, 2026, 02:32 AM
pricingcached input$0.195→$0.111
qwen/qwen3.5-397b-a17bupdateab879c8Jun 18, 2026, 08:32 AM
context_window262144→256000
pricinginput$0.39→$0.385output$2.34→$2.45
qwen/qwen3.5-397b-a17bupdate1dc75bcApr 17, 2026, 03:24 PM
pricingcached input$0.195
qwen/qwen3.5-397b-a17bupdate1cd659eMar 24, 2026, 04:19 AM
capabilitiesjson modeyes
qwen/qwen3.5-397b-a17bupdate2b100acMar 23, 2026, 05:30 AM
model_typechat
parameters397
active_parameters17
release_date2026-02-16
taglineThe Qwen3.5 series 397B-A17B native vision-language model...
tools["function_calling"]
qwen/qwen3.5-397b-a17bcreate7f5c66fMar 21, 2026, 05:16 AM