Opera introduces WebAI, a proposed web standard to run AI models locally in the browser

Opera has announced the rollout of WebAI, a feature that allows websites and browser extensions to run AI models directly within the browser, without relying on cloud servers. Positioned as a proposed web standard backed by major tech companies including Google and Microsoft, WebAI represents a shift toward decentralized, privacy-focused artificial intelligence. According to Krystian Kolondra, Executive Vice President of Browsers and AI at Opera, “A lot of what people want from AI doesn't need a data center. Your laptop can already do it, without sending your data anywhere.” By eliminating the need for separate AI apps, local server configurations, or API keys, Opera aims to democratize access to AI-powered capabilities while keeping user data local.

The early preview is now live across Opera One, Opera GX, and Opera Neon, supporting Windows PCs (64-bit, Intel/AMD) and Apple Silicon Macs, though Intel-based Macs, Linux, and Windows on ARM remain unsupported for now. Developers can leverage open-weight models through Opera's inference engine to build innovative extensions like Project Lens (which highlights article relevance), Got It? (reading comprehension quizzes), and Semantic Art (transforming articles into visual experiences). As Opera's Product Advocacy Director Dan Berte noted, this move empowers users to “start creating your own browser capabilities instead of waiting for software companies to build them”.

However, the local nature of WebAI comes with performance considerations. Processing occurs on the user's device using their own computing resources, meaning response times vary based on hardware capabilities. Opera acknowledges that “local AI performance is still improving”, but commits to ongoing optimization. The company first introduced local AI models to developers in 2024, and this expansion marks their broader push to make local inference a central component of their browser ecosystem.