Microsoft Edge Empowers Web Apps with New AI APIs
Microsoft is launching new APIs for its Edge browser, enabling developers to integrate AI capabilities directly into web applications. These APIs leverage built-in AI models, including Phi 4 mini, offering similar functionality to Google Chrome's built-in AI features.
On-Device AI for Enhanced Privacy and Performance
These new APIs provide websites and Edge extensions access to Phi 4 mini, a compact yet powerful AI model released by Microsoft. This model excels at mathematical computations thanks to its unique training data set. Its smaller size ensures efficient performance, especially on less powerful devices like laptops and mobile phones.
Beyond Phi 4 mini, Edge is also gaining additional AI APIs focused on writing assistance. These APIs will empower developers to create tools for generating, summarizing, and editing text. A translation API is also planned for release in the coming months, facilitating AI-powered language translations directly within Edge.
These experimental APIs are designed as potential web standards. They will function across platforms, browsers, and with other AI models. For developers working with sensitive data or in regulated industries, these APIs offer the privacy and security of on-device processing, eliminating the need to send data to external cloud services.
The new AI APIs are currently available for testing in the Edge Canary and Dev channels.
PDF Translation Arriving Soon
In related news, Microsoft is introducing a PDF translation tool to Edge. Supporting over 70 languages, this tool allows users to translate PDFs directly within the browser. Users can simply open a PDF in Edge, click the "Translate" icon in the address bar, and generate a translated document in their preferred language.
This PDF translation feature will be generally available next month, with Canary users gaining early access starting today.
