Opera Introduces WebAI to Run AI Models Locally in Its Browser

TL;DR

• Opera has introduced WebAI, a proposed web standard for running AI models directly in the browser.
• An early preview is available in Opera One, with support planned for Opera GX and Opera Neon.
• WebAI supports different models, including Google’s Gemma and Apple’s on-device foundation model on Macs.
• Local processing can reduce reliance on cloud services and keep information handled by the model on the user’s device.

Opera Brings Local AI to Its Flagship Browser

Opera has introduced WebAI, a new browser technology designed to let users run artificial intelligence models on their computers without relying on cloud processing.

The company is offering an early preview in Opera One, its flagship browser. Opera GX, which targets gamers and streamers, and the AI-focused Opera Neon are expected to receive the functionality later.

Users must opt in to enable the feature, according to Opera.

WebAI builds on the company’s earlier efforts to bring locally hosted language models into its browsers. Opera presents it as a web standard intended to support a broad range of models rather than depend on a single provider.

WebAI Supports Gemma and Apple’s On-Device Model

Opera named Google’s Gemma family as one example of the open-weight models WebAI can support. On Macs, the technology also supports Apple’s on-device foundation model.

The company says WebAI can handle tasks such as generating text, classifying content and identifying relationships between pieces of information directly within the browser.

Open-weight models make their underlying weights publicly available, giving developers greater flexibility to adapt them. Many are free to download and use, although their licenses can impose restrictions. They are therefore distinct from fully open-source software.

Opera says its approach is designed to accommodate a variety of these systems as the technology develops.

Local Processing Offers Privacy and Cost Benefits

Opera is emphasizing privacy as a central benefit of running AI locally.

Information processed by an on-device model does not need to be sent to a cloud AI provider for that processing. Users can also access supported features without purchasing a separate cloud-based AI subscription.

For developers, the approach could reduce the expense of adding AI capabilities to websites and applications.

“Developers can add AI features without running AI infrastructure or paying for every request,” Opera said.

The company added that users retain control over the AI running on their computers and that information handled by the local model stays on their devices.

That proposition comes as increasingly capable AI assistants gain access to more personal information, raising questions about how providers store and use it. Local processing gives users greater control over where AI computation happens, although it does not eliminate every privacy or security risk.

Hardware Requirements Remain a Trade-Off

Running models locally shifts the computing workload from a provider’s servers to the user’s device.

That can affect overall computer performance, particularly when a model requires substantial memory or processing power. Opera acknowledges that results will depend on the hardware available.

The preview will give users and developers an opportunity to assess those trade-offs as Opera expands local AI support across its browser lineup.