Opera launches local AI in browser: what it is and what to do about it
Opera has implemented the WebAI standard to launch local AI models in the browser: how it works, the benefits for privacy and business, and what small businesses should do.
One of the largest browser companies has just moved an experimental feature from development to release. We're talking about WebAI, a standard that allows AI models to run directly on a user's device without exchanging data with servers. The first functions are already working in Opera One, and will soon be available in Opera GX and Opera Neon. According to Habr, this is not just a set of features, but the beginning of browsers' transition to the role of local AI platforms.
For small business owners in Russia and the CIS, this is not an abstract technological news story. Local AI in the browser affects three main areas: web development (your websites and widgets), marketing (user behavior analysis), and security (data privacy). While most people are thinking about “cloud” AI, the first real alternative path has already come to the browser that millions use daily.
What is WebAI and how does it work on PC?
WebAI is a web standard that defines how a browser interacts with AI models. It is supported by Google, Microsoft and other players. The essence is that the code of a website or extension can request AI capabilities not through closed APIs of third-party services, but through the browser itself, where the model is already loaded locally.
Opera implements this standard based on open models. The user chooses the AI provider — at the start, models from Google and Apple are available. The browser downloads the model to the device and runs it locally. This means that the information being processed (texts, images, data from tabs) remains on the PC and does not go to the cloud. No subscription, no payment for requests, no account.
The capabilities that are already available look like basic but useful tools: sorting tabs by relevance, automatic highlighting of clickbait, searching for relevant jobs. All this happens without user involvement in the setup — you just need to enable the function.
Pros and cons: what local AI gives and where the trick is
| Advantages | Limitations |
|---|---|
| Data stays on the device, no leakage to the cloud | Performance depends on the user's hardware |
| No subscriptions and payment for each AI request | Models require sufficient resources (RAM, video memory) |
| Works without internet, after model download | Available only for 64-bit Windows and Apple Silicon Mac |
| The user controls permissions for each site | Pre-release version, functionality is limited |
| Open models, choice of provider | Quality of result is lower than that of large cloud models |
The main trick is that local AI is not a replacement for “cloud” giants. It solves narrow tasks: classification, sorting, generation of short texts. For complex analytical or creative tasks, it is not yet suitable. And performance really depends critically on the power of the PC. If the user has a weak laptop, the model may work slowly or not start at all.
What does this mean for web development and websites?
For developers, WebAI opens up the path to creating applications that do not depend on a backend. Imagine: you are making a website for a small store and adding a chatbot that answers questions about products. Previously, this required renting a server, setting up APIs, and paying for each request. Now you can upload a model to the user's browser, and it will work locally.
This is especially important for businesses in the context of sanctions and unstable access to foreign services. Local AI does not require a stable internet connection to work with the model. After the first download, everything happens offline. This reduces dependence on Western APIs that may be unavailable or blocked.
Opera plans to expand the number of open models optimized for different devices. In the future — tools to simplify the creation of browser applications based on local AI. For now, it's an experiment, but the foundation is being laid.
For marketing and analytics: new opportunities and risks
Marketers can use local AI to analyze user behavior on a website without transmitting personal data outside. For example, a browser extension can determine which products interested the visitor, what they clicked on, which pages they closed — and based on this generate relevant recommendations. All this happens on the device, without an account and without third-party tracking.
This is a big plus in terms of privacy. In the context of tightening legislation on personal data and growing security requirements, local processing is a way to avoid fines and loss of customer trust. The user himself decides which sites can access the AI and can revoke permission at any time.
However, there are risks. If AI works locally, then the analysis is limited to what the browser sees. It is impossible to obtain data from outside, for example, about the user's behavior on other sites. This means that marketing initiatives based on local AI will be less accurate than when using cloud analytics platforms.
What small businesses in Russia and the CIS should do: concrete steps
First — don't chase after fashion, but assess where local AI can provide real benefit. If you have a website, pay attention to web developers who are already experimenting with WebAI. Ask if it is possible to add simple functions to the site: a chatbot to answer frequently asked questions, a recommendation system, automatic sorting of goods. All this can be implemented without a backend, using a local model.
Second — check if your browser works with WebAI. Update Opera One to the latest version, enable the function in the settings. Try how it works: open several tabs, see if the browser sorts them by relevance. This will show what experience users can expect.
Third — think about privacy. If you collect customer data, local AI can be a way to protect it. Discuss with the developer how to implement processing of personal data on the user's device, without sending it to a server. This can be a competitive advantage, especially if you work with clients from Europe or other regions with strict data protection laws.
Fourth — keep an eye on the development of the standard. WebAI is still in development, but it is already supported by major companies. If you plan to scale your business and enter new markets, it is important to be aware of which tools are becoming standard. One way is to regularly read industry publications, such as Habr, and follow updates from Opera, Google, and Microsoft.
If you need help promoting a website where you decide to use AI tools, pay attention to digital marketing. Our specialists will help you set up contextual advertising and SEO so that you attract a target audience interested in technological solutions.
Frequently asked questions
How does local AI affect data security?
Data is processed on the user's device and not transmitted to the cloud. This reduces the risk of leakage and makes it easier to comply with personal data laws. However, the user himself must control which sites have access to the AI. The browser provides this capability through permissions.
Do I need to pay for using WebAI?
No. Models are downloaded and run locally. You only need to pay for the browser itself, which is free. Payment is required only if you use cloud AI services, but WebAI eliminates this.
Which devices support WebAI?
The pre-release version is available for 64-bit PCs running Windows (Intel, AMD) and Mac computers with Apple Silicon chips. Performance depends on the hardware. If the user has a weak laptop, model operation may be slow.
Can local AI replace cloud services for businesses?
Not yet. Local AI solves narrow tasks: sorting, classification, generation of short texts. For complex analytical and creative tasks, it is not suitable. Cloud models remain more powerful and accurate. But for specific scenarios where privacy and autonomy are important, local AI is a real alternative.
Web development, digital advertising and marketing expert since 2012. Founder of AIFORSALE.