Microsoft Foundry on Windows

A unified, reliable and secure platform supporting the AI developer lifecycle from model selection, fine-tuning, optimizing and deployment across CPU, GPU, NPU and cloud.

Windows AI APIs
APIs that unlock the power of models shipping with Windows.
Learn more
Foundry Local
Pull in your favorite open-source models and make your app smarter with Foundry Local
Learn more
Windows ML
Bring your own model and let Windows ML run it on the NPU, GPU or CPU.
Learn more
Tools and samples
Tools and samples get you started with Microsoft Foundry on Windows.
Learn more

IntegrateWindows AI APIs

Ready-to-use AI APIs powered by on-device models. Optimized for efficiency and performance on Copilot+ PCs devices with minimal setup required.

Learn more about Windows AI APIs

Phi Silica
Small yet powerful language model to generate text
Text Recognition (OCR)
Recognize text in images.
Image Super Resolution
Upscale your images using local AI models
Image Segmentation
Identify specific objects in an image
Image Description
Generate various types of text descriptions for an image
Object Erase
Retouch your images and remove unwanted objects

Customize Windows inbox models with your own data

Use LoRA for Phi Silica to customize the on-device small language model with your own data.
Build your vector store and bring semantic and lexical search in your apps using Semantic Search and Knowledge Retrieval.

Learn more

Deploy open-source models withFoundry Local

Find and integrate popular open-source language models with the Foundry Local CLI shipping on every Windows device.

Get started with Foundry Local

Microsoft Foundry on Windows integrates Foundry Local, offering developers quick access to ready-to-use open-source models on Windows Silicon. This offers the ability for developers to browse, test, interact and deploy models in their local apps. With the Foundry Local model catalog, we have done the heavy-lifting of optimizing these models across CPUs, GPUs and NPUs, making them ready-to-use.

Run your model on every device withWindows ML

Windows ML serves as the core of our AI platform and the integrated AI inferencing runtime on Windows. It allows developers to use their own models and deploy them efficiently across a broad silicon partner ecosystem, including AMD, Intel, NVIDIA, and Qualcomm, supporting CPU, GPU, and NPU architectures.

Learn more about Windows ML