AWS releases Strands Decider 2B open-source decision model

Amazon Web Services has released Strands Decider 2B, an open-source decision model built by its Strands Labs team. Unlike large language models, decision models skip text generation entirely, instead selecting from predefined choices and assigning confidence scores to each response. This approach delivers under 150 milliseconds of latency on local machines. Built on a modified Qwen3.5-2B base with a custom pointer head of just 1 million parameters, the model targets agentic tasks such as tool selection, model routing, and policy classification. It is available now via Hugging Face and GitHub.