Qwen: Qwen3 235B A22B Instruct 2507
qwen/qwen3-235b-a22b-2507
Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following,...
Modalities
Text
In / out price
₩158 / ₩965
1M
Context
262.14K
Released
Jul 22, 2025
Providers
Several places host the same model. Requests are spread by priority and health, and a local GPU always goes first when one is available.
| Provider | Context | Quantization | Max output | Priority | State |
|---|---|---|---|---|---|
| OpenRouter | 262.14K | — | — | 50 | unknown |