Opera Releases WebAI, a Standard for Running AI Models Inside the Browser

The early preview in Opera One runs open-weight models on your own computer, so developers can add AI features without paying for every request.

Opera has introduced WebAI, a new web standard for running AI models directly in the browser instead of on cloud servers, according to TechCrunch. It is available now as an early preview in Opera One, and users must opt in. Opera says Opera GX and Opera Neon will get it soon.

The standard is not tied to one model. It supports open-weight models (ones whose weights are published, so anyone can download and use them), and Opera points to Google's Gemma as an example. On Macs it also supports Apple's on-device foundation model. Opera says WebAI can generate text, classify content and work out how pieces of information relate, all inside the browser.

Opera's pitch to developers is that they can "add AI features without running AI infrastructure or paying for every request." Keeping processing on the device also means personal data does not have to go to a company's servers, though TechCrunch notes this does not remove every privacy risk.

There are trade-offs. Local models need more from your computer, and Opera says performance depends on the device. Open-weight models also come with license restrictions, and TechCrunch's report is based on Opera's announcement, so the company's claims are not independently verified.