Tongyi-Zhiwen/QwenLong-L1-32B

QwenLong-L1-32B is a 32 billion parameter long-context large reasoning model developed by Tongyi Lab, Alibaba Group. It is the first long-context LRM trained with reinforcement learning (RL) for enhanced long-context reasoning capabilities. The model excels in document question answering (DocQA) benchmarks, outperforming other flagship LRMs and achieving performance comparable to Claude-3.7-Sonnet-Thinking. It is optimized for robust long-context generalization across mathematical, logical, and multi-hop reasoning tasks.