Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a "thinking" mode for...
Specification
most-agreed values
Context
131K
Max output
16K
Released
—
Knowledge cutoff
2025-03-31
Retires
—
Open weights
—
Input
text
Output
text
Price
US dollars per million tokens · most-agreed
Input
$0.08
Output
$0.28
Available from 1 host
| Host | In $/M | Out $/M | Cache rd | Cache wr | Context | Output | Retires |
|---|---|---|---|---|---|---|---|
| openrouter | $0.08 | $0.28 | — | — | 131K | 16K | — |
Other listings of this model
this page is one of them
The corpus files Qwen: Qwen3 32B under several keys. This page is the qwen3-32b-04-28 listing. The full record — every host, every price and every benchmark score — is on the main Qwen: Qwen3 32B page.
More from Alibaba
most-hosted first