Amazon Web Services has released Strands Decider 2B, an open-source decision model inspired by TypeSafe's Jev. It arrived the same week OpenAI announced a similar offering, and is designed to sort among predefined options and return a confidence score. The model is small enough to run locally and is available now through Strands Labs.
The project began as a homebrew effort by AWS distinguished engineer Marc Brooker after he saw Jev. It briefly topped the Jevbench ranking for models of its size, and AWS engineers cleaned it up for release. Brooker said customer agentic workflows don't always need a full LLM, and a structured decider can be more reliable, lower latency, and potentially lower cost.
Like other decision models, Strands Decider is built on the "torso" of an LLM—here Qen3.5-2B—but outputs calibrated choices rather than generated text. Brooker sees a balance between speed and preserving general knowledge and language understanding. He also notes that smaller markets mean building something interesting can cost hundreds or thousands of dollars, so frontier labs may not dominate.
TypeSafe CEO Diogo Almeida is unconvinced by the flood of clones. "I get that people think it's a gold rush, but they might be underestimating the difficulty of making the models actually smart," he said, adding that the current batch looks more like ML enthusiasts implementing a cool architecture than a team dedicated to making intelligence useful. The source notes dozens of similar models have appeared since Jev debuted, raising questions about their long-term value.