Qwen3 VL 235B A22B Instruct

Qwentext + image → textHugging Face OpenRouter

The cheapest provider for Qwen3 VL 235B A22B Instruct is DeepInfra at $0.37 per 1M tokens (blended), charging $0.20 per 1M input and $0.88 per 1M output tokens. 5 providers serve it; the most expensive charges 1.71× as much. Context window: 256K tokens. Prices tracked since Oct 3, 2026.

Best price

$0.37

at DeepInfra

Input / output

$0.20 / $0.88

per 1M tokens

Providers

1.71× spread

7 days

—

best price

30 days

—

best price

Context

256K

tokens

Providers

Active offers ranked by blended price. Bars are relative to the most expensive.

Provider prices for Qwen: Qwen3 VL 235B A22B Instruct, USD per 1M tokens
Provider	Input	Output	Blended	vs cheapest	Context	Uptime 1d	Price since
DeepInfraCheapest fp8	$0.20	$0.88	$0.37	best	256K	95.2%	Oct 3
Alibaba	$0.26	$1.04	$0.455	+23%	128K	99.0%	Oct 3
Novita bf16	$0.30	$1.50	$0.60	+62%	128K	95.2%	Oct 3
Parasail fp8	$0.21	$1.90	$0.633	+71%	128K	99.6%	Oct 3
Venice fp8	$0.21	$1.90	$0.633	+71%	125K	95.0%	Oct 3

DeepInfraCheapest
$0.37
$0.20 in · $0.88 out
best
256K ctx95.2% uptimefp8
Alibaba
$0.455
$0.26 in · $1.04 out
+23%
128K ctx99.0% uptime
Novita
$0.60
$0.30 in · $1.50 out
+62%
128K ctx95.2% uptimebf16
Parasail
$0.633
$0.21 in · $1.90 out
+71%
128K ctx99.6% uptimefp8
Venice
$0.633
$0.21 in · $1.90 out
+71%
125K ctx95.0% uptimefp8

Blended = (3 × input + output) / 4.

Price history

Hover to compare providers at any point in time. Click a legend item to hide it.

USD per 1M tokens · each step is a price change

Not enough history yet

Tracking since Oct 3, 2026. The chart fills in as new crawls arrive every 6 hours.

Other pricing

Cache, reasoning and per-call fees. Token rates per 1M; others per unit.

Offered by	Cache read
DeepInfra	$0.11
Parasail, Venice	$0.10

Change log

Every recorded change for this model since Oct 3, 2026.

No changes yet

Tracking since Oct 3, 2026. Price changes, new providers and delistings will show up here.

About this model

Qwen3-VL-235B-A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video. The Instruct model targets general vision-language use (VQA, document parsing, chart/table...

Released: Sep 23, 2025
First tracked: Oct 3, 2026
Canonical slug: qwen/qwen3-vl-235b-a22b-instruct
Weights: Qwen/Qwen3-VL-235B-A22B-Instruct