Microsoft announced the launch of new APIs for the Edge browser, allowing developers to integrate generative AI features into web applications using models already embedded in Edge itself. One of the key innovations is access to the Phi 4 mini model — a compact version of the model with three billion eight hundred million parameters, which, according to the company, handles mathematical tasks well thanks to diverse training data, including both human-created and synthetic.
Web developers can use these APIs to add query input fields to their applications, as well as tools for generating, editing, and summarizing text directly in the browser. It is expected that within a few months, a text translation API will also be available, which will work based on AI and allow language translations directly in Edge.
Microsoft emphasizes that all new APIs operate on the user’s device, enhancing privacy and security since data is not sent to the cloud. Currently, these features are available for testing in the Edge Canary and Dev channels, and in the future, the company plans to make them available to a wider range of users and other platforms.
Additionally, the Edge browser will feature a PDF translation tool. Over seventy languages are supported, and users can open a PDF in Edge, click on the “Translate” icon in the address bar, and receive a new document in the chosen language. The feature will become publicly available next month, but Edge Canary users can already take advantage of it.
