Opera Launches WebAI Standard for Local Browser AI Processing
Opera introduced WebAI, a new standard allowing websites and extensions to run AI models locally on user devices to enhance privacy and reduce cloud costs.
Opera has introduced WebAI, a proposed web standard that enables websites and browser extensions to run artificial intelligence models locally on a user's device. Now available as an opt-in early preview in the Opera One browser, with support for Opera GX and Opera Neon, the system allows for text generation and content classification without sending personal data to external servers or requiring AI subscriptions.
Opera designed the system to be model-agnostic, supporting open-weight models such as Google's Gemma and Apple's on-device foundation models. To demonstrate the capability, the company released six experimental extensions that can flag clickbait or convert articles into quizzes. Hardware requirements include Apple silicon for Mac users and 64-bit Intel or AMD systems for Windows; however, Linux, Intel-based Macs, and Windows on Arm are currently excluded.
Krystian Kolondra, EVP of Browsers and AI at Opera, stated that many AI needs do not require a data center and that users will maintain control over what runs on their machines. While the move reduces per-request API costs for developers, analysts have raised concerns regarding enterprise security. Pareekh Jain of Pareekh Consulting noted that local processing reduces but does not eliminate risks, while Tulika Sheel of Kadence International emphasized that enterprises still require controls over how local models and browser permissions are managed across organizations.
