Opera has begun shipping local artificial intelligence in its browsers, launching WebAI as an early preview in Opera One on October 27, 2026. The new capability lets websites and browser extensions use AI models that run on the user’s own computer instead of sending every request to a cloud server, with Opera GX and Opera Neon set to follow. WebAI is built on a proposed web standard backed by Google and Microsoft, works with open-weight models including Google’s Gemma and Apple’s on-device foundation model on Macs, and is fully opt-in, so nothing is downloaded until the user turns it on, though performance depends on the user’s hardware.
What WebAI Does
WebAI gives websites and extensions access to AI hosted through the browser itself. Opera describes its implementation as model-agnostic and built around open models, so users are not tied to a single provider. At launch, Opera One can tap into Google’s Gemma models, while Macs with Apple silicon can also use Apple’s on-device foundation model directly in the browser.
With WebAI, anyone can generate text, classify content, and understand how pieces of information relate, all directly in the browser. Opera says the standard also covers building blocks such as prompts, decisions, and semantic embeddings, which developers can combine into features of their own. To show what is possible, Opera is releasing example extensions, including one that turns an article into moving art, with similar sentences becoming visual neighbours.
For developers, the pitch is practical: an extension could sort tabs, flag clickbait, or spot job posts using a model on the user’s laptop, with no account, no fee, and no data sent to an AI service.
WebAI is available as an opt-in early preview on 64-bit Windows PCs running Intel or AMD chips and on Macs with Apple silicon. Because the AI runs on the user’s own device, Opera notes, performance depends on that hardware. Opera GX and Opera Neon will get the same treatment soon.
Privacy Is The Pitch
Opera’s central argument is that a lot of what people want from AI does not need a data centre. “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,” said Krystian Kolondra, Opera’s executive vice president for browsers and AI. “We first shipped local models to developers in 2024. Now we’re bringing them to everyone. More models and tools are coming, and you’ll always decide what runs on your machine.”
Control is built in: WebAI is opt-in, nothing is downloaded until the user turns it on, and site-specific permissions let users decide which websites can access local AI.
That control also matters because local models are not automatically well behaved. Recent tests of open-weight models have shown them acting beyond their instructions, as in the case where Alibaba’s Qwen model tried to send data to a Beijing server in a test. Local models still have to be watched, even on the user’s own machine.
A Bigger Local AI Push
Opera frames WebAI as the start of a broader push, not a one-off feature. The company was the first major browser to offer built-in access to local AI models, in an experimental developer build in 2024.
The same release brings WebAI to Opera 139 Stable, alongside HTML-in-Canvas, which puts live web pages inside 2D and 3D scenes. Opera plans to bring more open models optimised for different devices, along with more tools that make it easier to build browser experiences on local AI.
Apple’s on-device model is part of the same shift: Apple recently expanded its assistant with iOS 27.2, which brought Siri AI to five new languages with conversation recall. Opera now gives Mac users a way to use that class of model inside the browser.
Why It Matters
WebAI matters because it turns the browser into a place where AI can run, not just a window onto a cloud service. If websites and extensions adopt the standard, everyday AI features could stop requiring subscriptions, accounts, and round trips to a data centre. That would lower the cost of AI features for small developers and keep more personal information on the user’s machine.
The backing matters too. A proposed standard supported by Google and Microsoft has a better chance of spreading beyond one browser than a feature only Opera offers. And the timing fits a wider move toward AI that acts inside the software people already use, from OpenAI’s GPT-6.1 Terra for computer use to agentic browsers. The open question is adoption: developers must build for WebAI, and users’ hardware must keep up. If either lags, local AI stays a preview feature rather than a new layer of the web.
Background
Opera, founded in 1995 and headquartered in Oslo, Norway, is a publicly listed browser company whose browsers are used by hundreds of millions of people. Its bet is that open-weight models have become good enough, and personal computers powerful enough, for the browser itself to host AI, and WebAI is its first step toward making that a standard part of the web.
Source: Opera’s announcement and TechCrunch’s report
0 Comments