Browser maker Opera has introduced a new web standard that allows users to run AI models directly in the browser without relying on the cloud.
WebAI, the new standard, is designed to work with a variety of AI models. Opera’s name, for one, dropped Google’s Gemma, the open-weight model that Google first introduced two years ago. On the Mac, WebAI also supports Apple’s on-device foundation model, the company said.
But Opera says WebAI is not tied to any particular model and is built to support a wide range of open weight AI systems. WebAI evolved from Opera’s earlier efforts to bring locally run models into the browser.
Opera is making new AI features available in early preview in its flagship Opera One browser. The company says Opera GX, a browser built for gamers and streamers, and Opera Neon, a browser that uses AI agents to perform tasks, will soon receive the same treatment. The company says users will have to choose whether to opt-in to this feature.
The open weight model offers many benefits to users. For one, they are usually free to download and use. The weights of the underlying model are also exposed, giving developers more freedom to customize how it works. Although not completely open-source software, the open-weight model still has some limitations.
The company says there’s an added benefit: privacy. Users can take advantage of AI capabilities without sending personal information to a company’s servers or paying for a cloud-based AI subscription.
“Developers can add AI capabilities without running AI infrastructure or paying per request fees,” the company said in an announcement. “Users do not need an AI subscription, the information processed by local AI remains on their device, and they can continue to control the AI on their computers.”
According to the company, WebAI enables anyone to generate text, categorize content, understand the relevance of information, and more, all directly within the browser.
Privacy concerns are growing as AI assistants become more powerful and gain access to personal information. Users often use AI chatbots to hand over large amounts of personal information, sometimes without a clear understanding of how that data will be stored or used.
Locally run models offer an alternative because the processing occurs on the user’s own device rather than on a company’s cloud server. This doesn’t eliminate all privacy risks, but it does give users a little more control. So it’s no surprise that Opera has made digital privacy one of its key selling points for years.
However, there are trade-offs. Running an AI model locally increases demands on your computer and can impact performance. Opera says the model’s performance depends on the device’s hardware.
If you make a purchase through links in our articles, we may earn a small commission. This does not affect editorial independence.
