Qwen3 235B A22B Instruct 2507

Qwen3-235B-A22B-Instruct-2507 is a 235B parameter model optimized for broad general capabilities, including logical reasoning, instruction following, mathematics, science, coding, and tool use. It delivers significantly improved comprehension and performance across multiple languages, with stronger alignment to user intent in subjective and open-ended tasks, resulting in more helpful and higher-quality text generation. The model also supports extended context windows up to 256K tokens, enabling advanced long-context reasoning and analysis.