Microsoft officially releases Foundry Local, which lets users distribute local AI environments via an installer. It also supports Mac and Linux

Microsoft has officially launched Foundry Local, a local AI runtime that developers can bundle and ship with their apps.

It provides a unified API for running inference on-device with multiple models, built on ONNX Runtime and Windows ML.

Foundry Local is optimized for GPU, NPU, and CPU, and now extends beyond Windows to Mac and Linux.

The move lets apps deliver AI features without cloud dependency or extra setup, improving privacy, availability, and deployment simplicity.