Continuing our coverage of key announcements from the Build 2025 developer conference—which, as expected, has placed a heavy emphasis on artificial intelligence—Microsoft unveiled a new platform called **Windows AI Foundry**, built on the **Windows Copilot Runtime**. This platform provides developers with a unified environment for creating, customizing, and deploying AI solutions within the Windows operating system.

Microsoft representatives noted that many developers need ready-made AI tools to handle specific tasks without having to build or train models from scratch. They also require a simple way to discover, test, and use open-source AI models.

**Windows AI Foundry** aims to address these needs by streamlining access to AI models through the integration of multiple model catalogs into a single platform. It supports the entire AI development lifecycle, including model selection, optimization, fine-tuning, and deployment. The platform connects to various model sources, such as **Foundry Local**, **Ollama**, and **NVIDIA NIMs (NVIDIA Inference Microservices)**.

Windows AI Foundry offers developers a wealth of opportunities, meeting them at every stage of their AI journey. It provides ready-to-use APIs powered by built-in models, tools for customizing Windows’ native models, and a high-performance runtime environment to help developers leverage their own models and deploy them across diverse hardware platforms. Thanks to the integration of **Foundry Local** into Windows AI Foundry, developers also gain access to a rich catalog of open-source models.

Pavan Davuluri, Corporate Vice President of Microsoft for Windows and Devices.

Windows AI Foundry includes **Windows ML**—a runtime environment that supports the deployment of custom models across a wide range of hardware, including CPUs, GPUs, and NPUs from leading manufacturers. Integrated model catalogs like **Foundry Local**, **Ollama**, and **NVIDIA NIMs** give developers easy access to open-source models for testing and local deployment.

Additionally, Windows AI Foundry offers out-of-the-box APIs for tasks such as smart data analysis, image recognition, and object removal. Advanced features have also been announced, including **LoRA-based fine-tuning** and APIs for **semantic search** and **knowledge retrieval**, enabling the creation of smarter, more personalized applications.

The **Windows AI Foundry** platform is now available to developers. For more details, visit the Windows Developer Blog.

Source