Opera's WebAI Standard Runs AI Models Locally in the Browser

Opera has shipped a new web standard, WebAI, that runs AI models directly in the browser — no cloud, no subscription, no data leaving the device. The Norwegian browser maker announced the standard as an early preview in its flagship Opera One browser, with Opera GX and Opera Neon slated to follow.

The standard is model-agnostic by design. Opera name-drops Google's Gemma — the open-weight model Google introduced two years ago — as one example. On Macs, WebAI also supports Apple's on-device foundation model. But the company says the standard isn't tied to any particular system and is built to handle a wide range of open-weight AI models.

Users have to opt in. Nothing runs by default, according to the company.

What does local AI actually buy you?

Open-weight models carry two practical advantages. They're typically free to download and use, and their underlying model weights are publicly available, which gives developers more freedom to customize how they work. There's a caveat: unlike fully open-source software, open-weight models still ship with some restrictions.

Then there's privacy. Because processing happens on the user's own device rather than a company's cloud server, users can tap AI features without sending personal information to corporate servers or paying for a cloud-based AI subscription.

Opera put the economics bluntly in its announcement:

"Developers can add AI features without running AI infrastructure or paying for every request. Users don't need AI subscriptions, the information the local AI handles stays on their own device, and they stay in control of the AI on their computer."

The standard lets anyone generate text, classify content, and understand how pieces of information relate — all directly in the browser.

In short: WebAI shifts both the cost and the data off the cloud and onto the user's machine.

Why privacy is the pitch

Privacy has become a growing concern as AI assistants get more powerful and gain access to more personal information. Users routinely hand over vast amounts of data to AI chatbots, sometimes without a clear picture of how that data is stored or used.

Locally run models offer an alternative. The processing stays on-device, which doesn't eliminate every privacy risk but does give users more control over their information. It's a natural fit for Opera, which has spent years making digital privacy one of its primary selling points.

The bet: users who don't trust cloud chatbots may still want AI features — they just want them on their own terms.

What's the catch?

Local AI isn't free in the hardware sense. Running models on-device puts more demand on the user's computer, which can affect performance. Opera itself notes that model performance will depend on the device's hardware — a heavier laptop will handle Gemma-class models better than a thin one.

WebAI didn't appear from nowhere. Opera says the standard grew out of its earlier work bringing locally run models into its browsers.

Where it rolls out next

The rollout sequence is clear:

• Opera One — available now as an early preview, opt-in
• Opera GX — the browser built for gamers and streamers, coming soon
• Opera Neon — the browser that uses AI agents to perform tasks, coming soon

If WebAI gains traction with developers, it could pressure rivals like Google and Microsoft to treat local, subscription-free AI as a browser feature rather than a cloud product.