Opera Brings AI Locally to Your Browser: Why ZDNET Calls It a Gamechanger

AI runs on your machine

Opera says its upcoming browser feature will use local AI models, so users’ queries and data remain on their machines. The company says open-weight models can also let developers build on them without costly AI infrastructure or paying for each request.

Many browsers rely on cloud-based AI responses, where user data may reach third parties and could be used to train large language models or build profiles. Opera says it wants to offer an alternative as models become smaller and hardware grows more powerful.

“A lot of what people want from AI doesn’t need a data center,” said Krystian Kolondra, Opera’s EVP Browsers and AI. He said the company first made local models available to developers in 2024 and is now bringing them to everyone.

How WebAI works

WebAI is a proposed web standard backed by companies including Google and Microsoft. It enables websites and extensions to use AI hosted in the browser. Opera says its WebAI implementation is model-agnostic and built around open-weight models. Users can select models from companies such as Google and Apple, and the feature is opt-in: downloads begin only when a user chooses to download a model.

Opera also plans to offer more open models optimized for particular devices, along with tools for building browser experiences that use local AI. The company says users will decide what runs on their machines.

Preview availability and limits

The early preview is available on 64-bit Windows PCs with Intel or AMD processors and on Macs with Apple silicon. It is not available on Intel-based Macs, Linux or Windows on Arm. Developers can explore example extensions described as using WebUI from “WebAI: Local AI for the web, now in Opera.” The preview gives users a way to try browser-hosted AI without making model downloads automatic.