Qwen vision-language model for visual reasoning, documents, and agent tasks
Specification
most-agreed values
Context
1M
Max output
131K
Released
2026-09-10
Knowledge cutoff
—
Retires
—
Open weights
no
Input
text, image, video
Output
text
Price
US dollars per million tokens · most-agreed
Input
$0.14
Output
$0.49
Cache read
$0.014
Available from 1 host
| Host | In $/M | Out $/M | Cache rd | Cache wr | Context | Output | Retires |
|---|---|---|---|---|---|---|---|
| venice | $0.14 | $0.49 | $0.014 | — | 1M | 131K | — |
Other listings of this model
this page is one of them
The corpus files Qwen 3.8 Flash under several keys. This page is the qwen-3-8-flash listing. The full record — every host, every price and every benchmark score — is on the main Qwen 3.8 Flash page.
More from Alibaba
most-hosted first