Brief
Amazon launches open‑source decision model Strands Decider 2B
AWS made the model publicly available after internal testing showed it topped the Jevbench ranking for its size.
By Felo News Desk · Published
Amazon Web Services released an open‑source decision model called Strands Decider 2B, according to TechCrunch. The model, built on the Qen3.5‑2B "torso" of a large language model, selects among predefined options and returns calibrated confidence scores rather than free‑form text.
What happened
The project began when Amazon distinguished engineer Marc Brooker adapted TypeSafe’s Jev model for internal use. After the home‑grown version briefly reached the top of the Jevbench ranking for models of its size, AWS engineers refined it and released it through Strands Labs, the unit developing AI‑agent tools.
What the report adds
TechCrunch notes that the release coincided with OpenAI’s announcement of a similar offering, highlighting a growing market for decision‑oriented models. The article explains that Strands Decider 2B is designed to be low‑latency, low‑cost, and small enough to run locally, addressing customer demand for workflow steps that do not require full‑scale language generation.
What was said
Brooker told TechCrunch, "What originally piqued my interest in this class of models was that they make a perfect decider for a workflow step—'what is the next thing for me to do here, based on where I am?'" He added that the model offers "a workflow step that can be structured in a way that is more reliable, thanks to the confidence scores, thanks to the closed domain of answers, [and is] lower latency, potentially lower cost."
How it came about
Brooker said the need emerged from conversations with AWS customers whose agentic workflows required a cheaper, faster alternative to full‑featured large language models. The model’s open‑source release follows Amazon’s broader push into AI tooling, complementing recent AWS announcements and the company’s ongoing development of Strands Labs projects.
Key facts
- Amazon Web Services released Strands Decider 2B, an open‑source decision model built on Qen3.5‑2B. (techcrunch.com)
- The model selects among predefined options and provides confidence scores instead of generating text. (techcrunch.com)
- Marc Brooker, an Amazon distinguished engineer, led the project after adapting TypeSafe’s Jev model. (techcrunch.com)
- Strands Decider 2B is positioned as low‑latency, low‑cost, and small enough to run locally. (techcrunch.com)
Sources
- [1] techcrunch.com — originally reported as “Amazon releases its own Jev clone as decision models flood the web”
