TL;DR
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Opera has introduced WebAI, a proposed browser standard that lets supported AI models run locally, with an early preview available in Opera One. The company says local processing can reduce the need to send data to cloud services, but performance depends on the user’s hardware and the preview’s capabilities and rollout remain limited.
Opera has introduced WebAI, a proposed web standard for running supported AI models directly in a browser on a user’s device, and has made an early preview available in Opera One. The company says the approach can let users access certain AI functions without sending the information those models process to a cloud server, though performance will depend on the device’s hardware.
Opera says WebAI is designed to work with a range of AI models rather than one model or provider. It identifies Google’s Gemma as an example of a supported open-weight model and says WebAI can also use Apple’s on-device foundation model on Macs. The company describes the standard as able to support a broader range of open-weight AI systems; the report does not provide a complete list of compatible models or supported devices.
The initial preview is in Opera One, Opera’s flagship browser. Opera says its gaming- and streaming-focused Opera GX and AI-agent browser Opera Neon will receive the features later. Users must opt in, according to the company. WebAI can be used for tasks including text generation, content classification and identifying relationships between pieces of information, Opera says.
The proposal grew out of Opera’s earlier work on running models locally in its browsers. Opera presents WebAI as a way for developers to add AI features without operating their own AI infrastructure or paying for each cloud request. These are company-stated advantages; the announcement does not provide independent performance tests, cost comparisons or details about the standard’s adoption by other browser makers.
Local AI Shifts Browser Data Handling
WebAI puts the browser forward as a place to run some AI tasks on a user’s own computer rather than sending prompts and related information to a remote service. That distinction matters for people wary of sharing personal details with online assistants, and for developers seeking to offer AI functions without routing every request through a paid cloud service.
Local execution is not a guarantee that all information stays private in every circumstance. The source report does not spell out how each model is downloaded, what telemetry the browser may collect, or how users can inspect and manage model activity. Still, keeping a task’s processing on-device can reduce the need to transmit its input to an AI provider, giving users a different option from cloud-based tools.
The approach also places responsibility on the user’s hardware. Opera cautions that model performance depends on the device, and running models can use computing resources that would otherwise be available to other applications. Users may see different speeds and capability depending on their computer and the model selected. Whether the convenience and privacy trade-offs work for them will depend on the task and hardware.
Products Worth Considering
Apple iPad Air 11-inch (M2), 512GB, Wi-Fi, Space Gray
Apple iPad Air (2025) 13‑inch Wi‑Fi 256 GB – Space Gray (M3 chip)
Amazon Kindle Scribe Colorsoft 64GB (1st Generation)
As an affiliate, we earn on qualifying purchases.
Opera Builds on Browser AI Work
WebAI follows Opera’s prior efforts to bring locally run models into its browsers. The company has also made digital privacy a prominent part of its browser positioning. This launch turns that broader direction into a named standard and an early product preview, but the supplied report does not describe a formal standards-body process or establish that other browser companies have adopted it.
Open-weight models have publicly available model weights, which can give developers greater freedom to adapt and run them. They are not necessarily fully open-source: licensing and other restrictions may still apply. Opera’s examples include Gemma and, on Macs, Apple’s on-device foundation model. The report does not give model versions, minimum hardware requirements, or a detailed account of what modifications WebAI permits.
“Developers can add AI features without running AI infrastructure or paying for every request.”
— Opera, in its announcement
Compatibility and Privacy Details Pending
The announcement, as described in the source report, does not specify when WebAI will leave preview, which models beyond the named examples are supported, or what hardware is required. It also gives no benchmarks for speed, memory use or model quality. Opera says results depend on device hardware, so users should not assume the same experience across computers.
It is also not clear whether WebAI will be adopted beyond Opera or how the proposal will be governed as a web standard. The company’s privacy claims concern information handled by local AI, but the report does not provide a technical audit or describe data collection by the browser itself. Local processing can reduce data sent to a model provider; it does not, by itself, establish that every aspect of a user’s activity stays on-device.
Opera Plans More Browser Previews
Opera says WebAI will come to Opera GX and Opera Neon, but it has not provided a release date in the source report. For now, the next step for users interested in the feature is to check the early preview in Opera One and confirm whether their device and selected model are supported.
Further details to watch for include model and hardware compatibility lists, performance information, privacy documentation and whether Opera opens the proposal to wider standards development. Until those are available, WebAI is best understood as an opt-in preview of Opera’s approach to browser-based, on-device AI—not evidence of broad industry adoption.
Key Questions
What is Opera WebAI?
WebAI is a proposed browser standard from Opera for running supported AI models on a user’s device through the browser, rather than relying entirely on cloud processing.
Which browser has the WebAI preview?
Opera says an early preview is available in Opera One. Opera GX and Opera Neon are expected to receive the features later, but no dates are given in the source report.
Which AI models does WebAI support?
Opera names Google’s Gemma as one example and says Apple’s on-device foundation model is supported on Macs. The report does not provide a complete compatibility list or model versions.
Does running AI locally guarantee privacy?
No. Local processing can mean that information handled by the model does not need to be sent to a cloud AI provider. The available report does not establish that all browser data stays on-device or provide an independent review of WebAI’s data handling.
Will WebAI work on every computer?
Opera says performance depends on a device’s hardware. The report does not specify minimum system requirements, so compatibility and speed may vary by computer and model.
Source: techcrunch.com
