Opera Neon: A Puzzling Experiment in the AI Browser Revolution
Opera’s latest creation, Neon, aims to redefine what an AI-powered browser can be. Marketed as a glimpse into the future of web navigation, it combines advanced artificial intelligence with Opera’s familiar browsing experience.
But while Neon promises a frictionless fusion between human and machine, what users actually get feels more like an overcomplicated experiment than a cohesive innovation.
For $19.90 a month, Neon presents not one, but three separate AI agents — each designed to perform unique tasks. The idea is bold: merge multiple AI capabilities into a single browser so users never need to leave the tab. The reality, however, is that juggling these agents quickly becomes confusing, leaving users wondering which AI to talk to — and why.
A Complex Architecture That Overcomplicates Simplicity
Neon isn’t just a regular browser with a built-in chatbot. It’s a three-in-one AI ecosystem operating inside a single interface
This structure should, in theory, provide flexibility and power. Instead, it often results in user fatigue and miscommunication between the human and the system
:Each AI serves a different function
- Chat: a conversational bot for questions and summaries
- Do: an active AI assistant capable of performing actions inside the browser
- Make: a creative agent for building or generating content
While the ambition is commendable, this division forces users to mentally switch between agents depending on the task — an unnecessary learning curve that slows down what’s supposed to be a smarter browsing experience.
A Crowded Market with a Steep Price Tag
Opera is stepping into a fiercely competitive field. Major players like Google Chrome (with Gemini), Perplexity’s Comet, and The Browser Company’s Dia are already offering AI-enhanced experiences — most of them free. By contrast, Opera’s decision to attach a nearly $20 monthly subscription makes Neon a premium gamble that demands perfection.
To its credit, Neon retains Opera’s best-known features, such as a built-in VPN, ad blocker, and a customizable sidebar with shortcuts to apps like WhatsApp and Messenger. But the spotlight clearly falls on its AI layer. On the new tab page, a toggle beneath the search bar lets users switch between traditional browsing and the three AI modes. Opera has confirmed that Neon uses technology from both OpenAI and Google, yet the company remains tight-lipped about which models power each agent leaving users guessing about what’s really behind the curtain.
Chat: Smart but Unreliable
Among the trio, Chat feels the most straightforward — a standard chatbot that can summarize pages or answer questions about what’s on-screen.
It handles simple queries well, like summarizing a recent paper on quantum computing or explaining a tech trend. However, its responses are often overly verbose, requiring users to trim down lengthy explanations manually.
In testing, Chat was asked to summarize reader comments from several Verge articles. It confidently produced a 400-word answer — all based on a false assumption that the articles had no comments.
In reality, they did. This revealed a serious limitation: Chat can’t always interpret or access dynamic web content, even when it appears to.
Opera’s EVP of Browsers, Christian Kolondra, later clarified that the issue stemmed from using the wrong tool a user should have asked the Do agent to expand the comment section first. Ironically, this explanation only highlighted the core issue: users have to know which AI does what, or they risk failure.
Do: The Overeager Assistant
“Do” is designed to take control — booking appointments, browsing shops, and performing real-time actions.
Unfortunately, it’s also where Neon becomes most unpredictable. During tests, Do tried to help find a flower bouquet but bizarrely added a funeral wreath to the cart.
In another case, it claimed no theater tickets were available for January, when a quick manual search proved otherwise.
The problem isn’t just inaccuracy — it’s the lack of control. When “Do” takes over, users can’t switch back to Chat in the same window or intervene mid-task. Only a faint red flash on the tab indicates something went wrong, signaling that human guidance is required.
Instead of simplifying web tasks, Do often creates new obstacles that users must fix themselves.
Even when it works, Do tends to move slower than manual browsing.
It demonstrates potential for automating routine searches but still relies too heavily on human correction far from the effortless autonomy Opera promises.
A Lesson in Complexity
Opera Neon captures the current identity crisis of AI browsers. The technology is powerful, yet the implementation feels disjointed.
Users must constantly decide whether to “Chat,” “Do,” or “Make,” turning simple online interactions into multi-step puzzles.
The irony is striking: a browser built to make the web smarter ends up making it more complicated. For nearly $20 a month, users expect seamless intelligence not a guessing game between three AIs.
Until companies like Opera can blend AI assistance into a truly unified system one that understands intent without demanding extra input AI browsers will remain fascinating prototypes rather than indispensable tools.
In short, Neon doesn’t fail because of what it can’t do, but because of how confusing it is to use.
Its steep subscription fee feels less like a payment for innovation and more like an entry ticket to an unfinished experiment in human-AI interaction.
