Amazon Web Services introduced Strands Decider 2B, a high-speed, low-cost model that sorts between pre-decided options and returns a confidence score for its choice. The model is fully open-sourced, available now, and small enough to run locally.
AWS distinguished engineer Marc Brooker started the project after seeing Jev and building his own version. That homebrew effort briefly reached the top spot on the Jevbench ranking for models of its size, prompting Amazon engineers to clean it up and release it through Strands Labs, an organization developing tools and protocols for deploying AI agents.
Brooker said the need emerged from conversations with AWS customers, whose agentic workflows did not always require the capability or cost of a fully featured LLM. "What originally piqued my interest in this class of models was that they make a perfect decider for a workflow step," he told TechCrunch, describing a system that offers lower latency and potentially lower cost while remaining reliable through confidence scores and a closed domain of answers.
Like other decision models, Strands Decider 2B is built on the "torso" of an LLM, in this case Qen3.5-2B, but instead of generating text it delivers calibrated choices. TypeSafe named its original model after economist William Stanley Jevons, invoking his theory that falling costs can increase demand.
Dozens of similar models have appeared since TypeSafe introduced the idea, raising questions about their long-term value. Brooker said the challenge is balancing speed and accuracy without degrading language understanding or general knowledge. He also said he does not expect frontier labs to dominate the space, since building something interesting can cost only hundreds or thousands of dollars.
TypeSafe executives said they are keeping their heads down and improving future models. "I get that people think it's a gold rush, but they might be underestimating the difficulty of making the models actually smart," the company's CEO said.