- Windows AI Foundry is Microsoft's new unified platform for developing and deploying AI models on local devices and in the cloud.
- Foundry Local enables developers to run, optimize, and manage AI models automatically adapting to available hardware, on both Windows and macOS.
- Integration with Windows ML and collaborations with partners such as Intel, AMD, Nvidia, and Qualcomm ensure efficiency and compatibility across devices.
- New APIs and development tools simplify the integration of AI into applications, enabling everything from model customization to advanced features like semantic search.
Microsoft has reaffirmed its commitment to artificial intelligence by presentar Windows AI Foundry and the service Foundry Local during its Build 2025 event, thus consolidating its strategy to place local AI at the center of the development experience on Windows and beyond. These new initiatives come with the goal of offering developers a A much more flexible, efficient and adapted environment to create, test and deploy AI models locally., eliminating many of the traditional barriers.
With the arrival of Windows AI Foundry, the company aims to facilitate the integration of AI models into applications Making the most of the hardware available on each device and promoting an open and robust ecosystem. This move represents a clear commitment to driving innovation at the user and business levels, Offering tools and APIs that simplify both building and customizing machine learning models for use offline or in the cloud.
Microsoft's commitment to local AI: What is Windows AI Foundry?
Windows AI Foundry It is the natural evolution of the previous Copilot Runtime, and It is presented as a centralized platform to manage the entire AI development lifecycleThis ranges from model selection and optimization to fine-tuning, inference, and integration into final applications, both on-premises and via the Azure cloud.
Developers can access open source AI models and proprietary models, including the most popular ones on the market such as Ollama, Nvidia NIM and other options that allow you to take advantage of different model catalogs. One of its strong points is its ability to automatically detect and adapt to local hardware (CPU, GPU, NPU), recommending and optimizing the most appropriate models in each case without the need to manually configure each environment.
La collaboration with leading hardware manufacturers Such as Intel, AMD, Nvidia, and Qualcomm have been key to ensuring that Windows AI Foundry runs smoothly, allowing developers to work with different chipsets and devices, from lightweight laptops to powerful workstations or even specialized hardware like Copilot+ PCs. Microsoft is responsible for keeping dependencies always up to date to ensure maximum compatibility and performance as processors evolve.
Foundry Local: AI without relying on the cloud
Foundry Local It stands out as a module specifically designed for those looking to run AI models completely locally. Thanks to its cross-platform design, developers can deploy models not only on Windows but also on macOS, facilitating the creation of hybrid applications or mixed work environments without sacrificing compatibility.
Foundry Local provides specific tools such as SDK and CLI to test, optimize, and deploy models directly into the user environment. You can view the list of models compatible with each computer's hardware and run or customize them as needed. Plus, take advantage of ONNX Runtime, Microsoft's open inference engine that accelerates the execution of AI models on different architectures.
This solution also includes support for LoRA (Low-Rank Adaptation), a technique that allows predefined models to be fitted to custom data sets quickly and efficiently, making it ideal for companies or organizations that want to adapt AI to specific cases without developing models from scratch.
APIs and new capabilities for developers
Microsoft has placed great emphasis on make AI adoption as easy as possible. For that, Windows AI Foundry brings a series of ready-to-use APIs focused on common tasks such as text intelligence, image recognition, object description, and semantic search. These tools eliminate the need to develop custom models for basic tasks, allowing developers to focus on unique features for their applications.
Another important advantage is the support of APIs for advanced search and knowledge retrieval features, specifically designed to be integrated into augmented retrieval (AR) workflows, facilitating the query and analysis of both structured and unstructured information.
All this is complemented by the integration of Windows ML, Microsoft's local inference runtime, which simplifies production deployment and eliminates the need to package runtimes or drivers separately. With automatic updates, compatibility and security are always guaranteed.
Compatibility, security, and a unified experience
The ecosystem proposed by Microsoft also relies on the adoption of open standards such as Model Context Protocol (MCP), so that AI agents and models can communicate seamlessly and securely across different tools and platforms. Microsoft is committed to true interoperability, facilitating the industry's ongoing evolution and integration with third-party solutions.
In terms of security, the platform integrates advanced features such as secure execution environments (with support for Virtualization Based Security and post-quantum cryptography), as well as observability tools, performance monitoring and threat protection.
Microsoft's platform with Windows AI Foundry and Foundry Local not only makes life easier for developers, but also expands the possibilities for businesses and power users looking to Harness the potential of AI without losing control or privacy.
Microsoft is advancing the future of AI development, promoting a model in which artificial intelligence can be run efficiently, easily and safely directly on users' devices, without relying exclusively on the cloud. The introduction of Windows AI Foundry and Foundry Local places Windows at the forefront of local AI services, offering a flexible, open platform fully adapted to the needs of the ecosystem and its technology partners.
I am a technology enthusiast who has turned his "geek" interests into a profession. I have spent more than 10 years of my life using cutting-edge technology and tinkering with all kinds of programs out of pure curiosity. Now I have specialized in computer technology and video games. This is because for more than 5 years I have been writing for various websites on technology and video games, creating articles that seek to give you the information you need in a language that is understandable to everyone.
If you have any questions, my knowledge ranges from everything related to the Windows operating system as well as Android for mobile phones. And my commitment is to you, I am always willing to spend a few minutes and help you resolve any questions you may have in this internet world.