TL;DR: Opera has launched Opera WebAI, an opt-in early preview that lets websites and browser extensions run AI models locally on your computer instead of in the cloud. It’s available now in Opera One on 64-bit Windows PCs (Intel/AMD) and Apple silicon Macs. WebAI adds Prompt, Decision and Semantic Embeddings APIs for developers, supports open-weight models such as Google’s Gemma (plus Apple’s on-device model on Macs), and needs no API key or AI subscription. Opera says WebAI is a proposed web standard backed by companies including Google and Microsoft.
On October 27, Opera announced what it calls “a bigger push into local AI.” The first step, Opera WebAI, turns the browser itself into a small AI engine that any website or extension can call, with your permission. The idea is simple: lots of everyday AI tasks don’t need a data center, and keeping the work on your own machine means your data doesn’t have to leave it. Here’s what WebAI does, how to try it and what the catches are.
What is Opera WebAI?
According to Opera’s announcement, WebAI “lets websites and extensions use AI that runs locally, in your browser.” Local inference is built into the browser, so there’s no separate AI app to install, no local server to set up and no API key to get.
Opera describes WebAI as a proposed web standard, backed by companies including Google and Microsoft. Its own implementation is model-agnostic: it can use open-weight models Opera provides and compatible models already on your device. TechCrunch reports that Opera names Google’s Gemma as one example and says WebAI also supports Apple’s on-device foundation model on Macs.
“A lot of what people want from AI doesn’t need a data center. Your laptop can already do it, without sending your data anywhere,” said Krystian Kolondra, Opera’s EVP of Browsers and AI, in the announcement. “We first shipped local models to developers in 2024. Now we’re bringing them to everyone.”
Opera WebAI at a glance
| Detail | What Opera says |
|---|---|
| Status | Early preview, opt-in |
| Browser | Opera One now; Opera GX and Opera Neon also named in the rollout |
| Supported systems | Windows PC (64-bit, Intel/AMD) and Macs with Apple silicon |
| Not supported | Intel-based Macs, Linux, Windows on Arm |
| Developer APIs | Prompt, Decision and Semantic Embeddings |
| Models | Open-weight models via Opera (Gemma named by TechCrunch), plus compatible on-device models such as Apple’s on Macs |
| Cost | No API key, AI subscription or per-request cloud fees |
One detail is worth flagging. Opera’s blog says WebAI is in early preview “across Opera One, Opera GX and Opera Neon,” but its how-to steps only cover Opera One. Digital Trends and TechCrunch both report that GX and Neon are due to get it later, so Opera One is the browser to use today.
What developers can do with WebAI
WebAI gives web developers three building blocks:
- Prompt API: generate text from a prompt, like a mini local chatbot.
- Decision API: classify information, for example sorting content into categories.
- Semantic Embeddings API: find relationships between pieces of content, the same idea behind semantic search. (New to the concept? See our explainer on what embeddings are in AI.)
Opera pitches two main benefits for developers, especially small teams: page data can stay on the user’s device when AI processes it, and there are no per-request cloud costs or AI infrastructure to run. Full documentation and hands-on playgrounds are at next.opera.com/web-ai.
Opera’s six demo extensions
To show what’s possible, Opera built six experimental extensions. It stresses these are “examples to spark ideas, not Opera features”:
- Project Lens: colors each paragraph of an article by how relevant it is to a project you describe.
- Got It?: writes a quick quiz about the article you just read.
- Questionable Intelligence: asks playful questions about an article and draws the verdict as a chart.
- Semantic Art: turns an article into moving art where similar sentences sit near each other.
- RecordIn Alchemy: turns any article into an “insufferably inspirational” social media post.
- C·I·A·O (Cat Intelligence at Opera): two ASCII cats improvise a scene on any topic you give them.
How to try Opera WebAI
Opera lists three steps:
- Download or update Opera One.
- Go to Settings → Opera AI and turn on Enable Web AI API. This downloads a compatible local model, so it may take a few minutes.
- Install one of the example extensions and open it from your extensions menu.
Nothing is downloaded until you switch WebAI on, and you can turn it off at any time. Digital Trends adds that each site needs your permission before it can use the browser’s local AI.
Why local AI in the browser matters
Most AI features today send your text to a company’s servers. That raises cost for developers and privacy questions for users. Running models on-device flips that: the content the AI handles doesn’t need to go to a cloud AI provider. We’ve seen the same trend elsewhere, such as Perplexity Computer running agent tasks fully on a Mac and people running image generators on their own PCs.
What makes WebAI different is that it opens local AI to any website or extension through standard browser APIs, rather than keeping it inside one app. As Opera puts it, you can “start creating your own browser capabilities instead of waiting for software companies to build them.”
The catches
- Performance depends on your hardware. Opera says response speed depends on your device, and local AI performance “is still improving.” Digital Trends notes Opera hasn’t published minimum performance requirements beyond the supported platforms.
- Limited platforms. No Linux, no Intel Macs and no Windows on Arm for now.
- Early preview. Expect rough edges, and remember the demo extensions are experiments.
- A proposed standard, not a finished one. Opera says the effort has backing from Google and Microsoft, but developers will want to watch whether other browsers ship compatible APIs before relying on it.
- Local isn’t risk-free. TechCrunch points out that on-device processing doesn’t eliminate every privacy risk, though it gives users more control.
Opera also used the same announcement to launch HTML-in-Canvas, which it says puts live web pages inside 3D worlds, according to its newsroom.
FAQ
Is Opera WebAI free?
Opera says there’s no AI subscription or API key needed; the model runs on your own computer.
Does WebAI send my data to the cloud?
According to Opera, because the AI runs on your computer, the content it processes doesn’t need to be sent to a cloud AI provider.
Which models does WebAI use?
Open-weight models provided through Opera, such as Google’s Gemma, and compatible models already on your device, including Apple’s on-device foundation model on supported Macs.
The Teck Hustlers Editorial Team covers technology, business, finance and gaming news for readers who want the story behind the headline. Every article is researched, written and edited by our in-house team, checked against company announcements, filings and reputable reporting, and updated when new facts emerge. We focus on what each development means for founders, professionals, investors and everyday users. Have a tip or a correction? Reach us through our contact page.
