Qwen3 235B A22B

Qwen3 represents the latest evolution in the Qwen series of large language models, featuring a combination of dense and mixture-of-experts (MoE) models. It introduces groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support. Key features include seamless switching between thinking mode (for complex reasoning, math, and coding) and non-thinking mode (for general-purpose dialogue), ensuring optimal performance across various scenarios. Qwen3 excels in mathematical and commonsense reasoning, surpassing previous Qwen models in both thinking and non-thinking modes. It also offers superior human preference alignment, especially in creative writing, role-playing, multi-turn dialogues, and instruction following. With strong capabilities in agent-based tasks and support for over 100 languages and dialects, Qwen3 is optimized for complex agent interactions and multilingual instruction following.