Skip to main content
Back to Explore

Qwen: Qwen3.5-Flash

Alibaba·Released 2026-02-25
Open Source1.0M ctxMultimodal

About

The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. Compared to the...

Pricing

Input

$0.07

per 1M tokens

Output

$0.26

per 1M tokens

Blended

$0.11

per 1M tokens

Cheaper than 82% of models. Median price is $0.67/1M tokens.

Cost Calculator

Tokens per day1M
100K100M

Daily

$0.11

Monthly

$3.41

vs. Similar Models

Qwen: Qwen3.5-9B
$0.11-1%
Meta: Llama 3.2 3B Instruct
$0.12+5%
Qwen: Qwen3 Coder 30B A3B Instruct
$0.12+5%
Granite 4.0 H Small
$0.11-6%

Performance

Context Window

1.0M

tokens

Larger than 80% of models

Max Output

66K

tokens

7% of context

Context Window Comparison

Anthropic: Claude Fable 5
1.0MSame
Anthropic: Claude Opus 4.8
1.0MSame
Anthropic: Claude Opus 4.7
1.0MSame

Open Source

Quick Compare

Similar Models

Compare all 7 models