ModelWatch
← Catalog

Qwen3.5-Flash

Alibaba (Qwen)qwenlm.github.io

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...

visionlong-contextlow-cost
Input price
$0.07 / 1M tokens
Output price
$0.26 / 1M tokens
Context window
1M tokens
Max output
66K tokens
Output speed
n/a
Time to first token
n/a
Inputs
text, image, video
Open weights
No
OpenRouter id
qwen/qwen3.5-flash-02-23
Listed
Feb 25, 2026

Benchmarks

Intelligence Index
n/a
Engineering Index
n/a

No independent speed or benchmark data for this model yet. Prices, context window and listing date are live from OpenRouter.

Compare with others