Qwen: Qwen3 235B A22B Instruct 2507
5.4 / 10
262.1K context
Proprietary
About this model
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,...
Benchmark Scores
SimpleQA
54.3
Technical Specs
- Architecture: text->text
- Context Window: 262,144 tokens
- Input Modalities: text
Hardware Requirements
- API-only (no local hardware needed)
Pricing
| Input | Output | Currency |
|---|---|---|
| 0.09 / 1M tokens | 0.55 / 1M tokens | USD |