Qwen: Qwen2.5 VL 72B Instruct

Name: Qwen: Qwen2.5 VL 72B Instruct
Author: OpenRouter

apache-2.072B

Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects.

Context

128K

Max output

115K

Input price

$0.8/1M tokens

Output price

$1/1M tokens

Capabilities

vision

tool call

structured output

reasoning

json mode

streaming

fine tuning

batch

Details

Model IDqwen/qwen2.5-vl-72b-instruct

ProviderOpenRouter

CreatorQwen

Familyqwen2.5

Licenseapache-2.0

Parameters72B

Status—

Input modalitiestext, image

Output modalitiestext

Architecture—

Knowledge cutoff—

Training data cutoff—

Release date2025-02-01

Deprecation date—

Typechat

Reasoning tokens—

Max input—

Open weightYes

Sourceofficial

Last updated2026-09-04

Pricing

Input

$0.8

↑220%$0.25

Output

↑33%$0.75

Cache write

—

Cache read

$0.4

Batch in

—

Batch out

—

Price history · $/1M input

$0.8

$0.25

$0.8

$0.25

$0.8

$0.25

$0.8

Mar 21Apr 17Jun 18Jul 12Jul 17Aug 5Sep 4

Date	Input	Output	Cache read	Change
Mar 21, 2026	$0.8	$0.8	—	—
Apr 17, 20261dc75bc	$0.25	$0.75	—	↓69%
Jun 18, 2026ab879c8	$0.8	$1	$0.4	↑220%
Jul 12, 2026bbbc783	$0.25	$0.75	$0.4	↓69%
Jul 17, 202620c73ac	$0.8	$1	$0.4	↑220%
Aug 5, 202619c9fe3	$0.25	$0.75	$0.4	↓69%
Sep 4, 20260d4ae28	$0.8	$1	$0.4	↑220%

Family Comparison: qwen2.5

Model	Caps	Perf	Speed	Context	Max out	Pricing
Qwen: Qwen2.5 Coder 7B Instruct	RVTS	—	—	33K	—	$0.03$0.09
Qwen: Qwen2.5 VL 32B Instruct	RVTS	—	—	128K	—	$0.2$0.6
Qwen: Qwen2.5 VL 72B Instruct	RVTS	—	—	128K	115K	$0.8$1

API

GET/v1/models/openrouter/qwen/qwen2.5-vl-72b-instruct

Qwen: Qwen2.5 VL 72B Instruct

qwen/qwen2.5-vl-72b-instruct

Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects.

Changes · 12 entries

qwen/qwen2.5-vl-72b-instructupdate0d4ae28Sep 4, 2026, 06:37 AM

max_output_tokens28800→115200

pricinginput$0.25→$0.8output$0.75→$1

qwen/qwen2.5-vl-72b-instructupdatee3d7dedAug 31, 2026, 04:13 AM

max_output_tokens128000→28800

qwen/qwen2.5-vl-72b-instructupdate19c9fe3Aug 5, 2026, 03:04 PM

pricinginput$0.8→$0.25output$1→$0.75

qwen/qwen2.5-vl-72b-instructupdatebf78dcfJul 25, 2026, 05:49 AM

context_window131072→128000

qwen/qwen2.5-vl-72b-instructupdate20c73acJul 17, 2026, 02:34 AM

pricinginput$0.25→$0.8output$0.75→$1

qwen/qwen2.5-vl-72b-instructupdatebbbc783Jul 12, 2026, 07:45 AM

pricinginput$0.8→$0.25output$1→$0.75

qwen/qwen2.5-vl-72b-instructupdateab879c8Jun 18, 2026, 08:32 AM

max_output_tokens32768→128000

pricinginput$0.25→$0.8output$0.75→$1cached input$0.4

qwen/qwen2.5-vl-72b-instructupdate73d11c6May 18, 2026, 03:39 AM

context_window32000→131072

qwen/qwen2.5-vl-72b-instructupdate1dc75bcApr 17, 2026, 03:24 PM

context_window32768→32000

pricinginput$0.8→$0.25output$0.8→$0.75

qwen/qwen2.5-vl-72b-instructupdate1cd659eMar 24, 2026, 04:19 AM

capabilitiesjson modeyes

qwen/qwen2.5-vl-72b-instructupdate2b100acMar 23, 2026, 05:30 AM

model_typechat

parameters72

release_date2025-02-01

taglineQwen2.5-VL is proficient in recognizing common objects su...

qwen/qwen2.5-vl-72b-instructcreate7f5c66fMar 21, 2026, 05:16 AM