Microsoft is rolling out a suite of new APIs in its Edge browser designed to enable web developers to integrate artificial intelligence features directly into their web applications. Announced at Build 2025, these offerings closely resemble Google Chrome’s built-in AI functions but represent Microsoft’s effort to compete for developer attention.
Among these updates, developers can now access the Edge AI APIs, Phi 4 mini AI model through both websites and Edge extensions. Microsoft highlights that the Phi 4 mini, released in late February, contains around 3.8 billion parameters and delivers strong performance on mathematical tasks owing to a diverse training dataset that blends human-crafted and synthetic inputs.
While models with more parameters tend to excel at problem solving, more compact models like Phi 4 mini can operate smoothly even on devices with limited processing power, such as smartphones and laptops. Microsoft intends for these APIs to make AI more accessible to a broad range of developers by focusing on efficiency as well as capability.
Edge Empowers Developers With Expanded AI Capabilities
In addition to general-purpose APIs, Microsoft is unveiling a set of writing assistance tools for automating the creation, summarization, and modification of content within web apps. On the horizon, the company plans to introduce a translation API that will support AI-driven language conversion natively within Edge.
According to Microsoft, these APIs are currently experimental but could lead to future web standards, offering cross-browser compatibility and integration with other AI models. Importantly, they promise on-device data processing, enhancing privacy and ensuring that sensitive information never leaves the local machine.
Presently, developers can experiment with these new AI features via the Edge Canary and Dev preview channels, giving early access to cutting-edge technology. This approach aims to accommodate organizations with heightened privacy requirements or those working in regulated environments.
In related updates, Microsoft has announced a translation feature for PDFs within Edge, supporting translations in over 70 languages. This tool allows users to open PDFs, use the new translate functionality in the Edge address bar, and quickly produce a new version of the document in their preferred language.
The PDF translation tool will become available to all users next month, while early adopters using the Canary version of Edge can try it immediately. This upgrade aligns with Microsoft’s broader vision for Edge AI APIs, Phi 4 mini as a hub for innovative productivity and AI-powered tools.
