Amazon Web Services has launched Strands Decider 2B, an open-source decision model designed for automated workflows. The project originated from Amazon engineer Marc Brooker, who developed it after observing the Jev model created by TypeSafe. The tool is now available through Strands Labs, an internal Amazon organization focused on AI agent deployment.
The model is based on the Qen3.5-2B architecture and is optimized for speed and low cost. Unlike general-purpose large language models, Strands Decider 2B is built to select from predefined options and provide confidence scores for its choices. It is small enough to be deployed locally, offering lower latency for specific workflow tasks.
The release follows a trend of developers seeking intelligence tailored for automation rather than generative tasks. While dozens of similar models have emerged, Amazon engineers aim to balance decision-making accuracy with the model's general knowledge. Development costs for such specialized tools are estimated in the hundreds or thousands of dollars.
TypeSafe CEO Diogo Almeida noted that while interest in the sector is growing, creating truly intelligent decision models remains a significant challenge. He stated that his company continues to focus on improving its own models and does not currently view the recent wave of similar implementations as direct competition.
