SOURCE-LINKED INTELLIGENCE
Qwen: Qwen3 235B A22B Instruct 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,...
Read original source ↗ Open in workspace
- recordType
- model-catalog
- evidenceStatus
- provider-reported
- region
- Global
Evidence & attribution
First collected: 2026-09-20T08:01:03.945Z. This is not the publication date.