Microsoft released its new ‘Microsoft-Decision-1.1’ decision model on Azure immediately following its unveiling at the annual Ignite conference by CEO Satya Nadella. The company will license the model’s weights, enabling customers to build customized versions of the system for their specific needs. This approach contrasts with the standard API-only model often used by competitors.
The new model was engineered to balance cost, latency, intelligence, and calibration according to Microsoft’s internal teams. While Microsoft’s offering enters the market, TypeSafe’s Jev family of decision models continues to lead usage statistics, having processed over 1.5 billion queries. In the startup ecosystem, Sqwish Labs announced a $3 million pre-seed round for its own decision model ‘kettei fat’ and offers tools to improve models with production data.
This move by Microsoft allows enterprises to fine-tune decision logic directly on their own infrastructure, which could reduce reliance on third-party providers. For developers working on complex automation, having a licensed decision model means more control over how the system processes inputs and generates outputs. The availability of customizable weights on Azure provides a direct alternative to open-source options for companies needing proprietary decision engines.



