Amazon Web Services has released Strands Decider 2B, a fully open-source decision model designed to help AI agents rapidly choose among predefined options and provide calibrated confidence scores. The model is small enough to run locally and targets workflow steps where a full-featured large language model may be unnecessarily costly or slow. Amazon distinguished engineer Marc Brooker developed the project after studying TypeSafe’s Jev, and an early version briefly topped Jevbench for models of its size before Amazon engineers refined it through Strands Labs. Built on the Qen3.5-2B LLM torso, Strands Decider produces choices rather than prose. Its release came the same week OpenAI announced a similar offering, underscoring growing interest in specialized decision models. Brooker said the main technical challenge is improving accuracy and calibration without weakening the model’s language understanding, knowledge and general-purpose usefulness. TypeSafe founder and CEO Diogo Almeida said the company does not yet see serious competition, arguing that making these models genuinely intelligent remains difficult.