Opera has introduced WebAI, a proposed web standard that lets websites and extensions run AI models directly in the browser, on the user’s own computer, instead of sending every request to the cloud. The feature is available as an opt-in early preview in Opera One, with Opera GX and Opera Neon set to follow.
WebAI is not tied to a single model. Opera names Google’s open-weight Gemma models as an example, and on Macs it also supports Apple’s on-device foundation model. The company says the standard is built to support a wide range of open-weight systems, and it covers text generation, content classification and measuring how pieces of information relate to each other. Opera has also published six experimental extensions built on WebAI, including a tab organiser and a clickbait flagger.
The pitch is straightforward: developers can add AI features without running their own AI infrastructure or paying per request, and users need no AI subscription, with the information handled by the local model staying on their own device. That matters for private or repetitive work, such as summarising a document you would rather not upload. The trade-off is hardware: running models locally puts more demand on the user’s computer, and performance will depend on the device. At launch the preview runs on 64-bit Windows PCs with Intel or AMD processors and on Macs with Apple silicon, but not on Intel-based Macs, Linux or Windows on Arm.
Local models are already familiar through tools such as Ollama, while cloud assistants such as ChatGPT and Perplexity remain the default for heavier tasks — see our Perplexity vs ChatGPT comparison.
WebAI is still an early preview, but it could shift everyday AI from the data centre into the browser tab itself.
Source: TechCrunch
