Remember Browser Operator? Opera is taking all of this agentic AI tech and launching it in a new browser called Opera Neon.
But there’s more to it than that. With Neon, the Norwegian browser company is keen to build a one-stop shop for basically any sort of AI interactions — be it chatting to an AI agent or even getting it to build and code games while you sleep.
The official launch date is unknown (as is the subscription price), but a waitlist is now open for sign ups and will let you try the Alpha when it launches. Let’s get into what it can do, and how it compares to the likes of what ChatGPT and Google Gemini can do.
What can it do?
Its functionalities are split into three different categories – Chat, Do and Make.
- Chat: Think of this as your standard ChatGPT interaction – a prompt box that you detail what you want in and the AI will go out and find out. There’s also a multi-modality to it as well, as it’s not just a plain text box. It can also provide context about the webpage you’re on, and search the web to tailor its answer with the most relevant context.
- Do: This is the big agentic part. Taking the tech I got to test in Browser Operator, it can browse the web on your behalf, fill in forms, book trips and even do shopping for you – all while “preserving your security and privacy.” I’ll go into more detail on this in a minute.
- Make: Easily the coolest addition in my books – built into Neon is an AI engine that is able to understand whatever you want to build or code, and build it for you. It can even do so offline via a cloud computer, so you can go do other things while Neon gets to work.
So on paper, it seems like Opera has taken this nugget of great tech in Browser Operator and turned it up to 11 by making a big bet on what the future of web browsing will actually be.
It’s a bold take, and after testing the underlying tech myself a month ago, I do believe that an agentic assistant is core to the future of web browsing. But there’s a couple of questions to overcome first.
How does it work?
Let’s dig into this a little deeper, as I am the only person outside of Opera to have played around with the Browser Operator that runs underneath a lot of this.
Up until now, agentic browsers have taken screenshots of a website, interpreted them with vision AI and decided the next step. This required a remote desktop session, and has been demonstrably slow and clumsy in what it can do.
Opera Neon is set to do things differently – stripping away those UI elements that cause agentic AI systems to trip over so often and focus on the “textual representation of websites” to interact with them.
In my time using Browser Operator and talking to the Opera team, this is taking advantage of something web developers call the “a11y” level – the accessibility reader mode that a lot of websites have, which just so happens to make it a whole lot easier for an agentic AI to browse too.
And the end result is a dramatic speed boost over the more clunkier way of doing it. Pair that with the AI agent that looks set to work similarly to what we see in the chatbot overlay on Opera One R2, and that builder too, and this is looking like a pretty sweet package.
But will people pay for it?
That’s the ultimate question. Opera Neon is a “premium subscription product” in the company’s own words, and the word “premium” is doing a lot of heavy lifting here.
With the way this has been built around three core pillars that the likes of a ChatGPT Plus subscription or Gemini Advanced (giving you Gemini AI in Chrome) would get you, Opera needs to either aim for $20 or undercut the competition for the chance to really get out ahead here.
Speaking to Opera, the team assured me that the browsing part “will remain free of charge,” and the subscription covers the “AI agentic capabilities of Opera Neon.” So your browser isn’t just going to stop you from surfing if you’re unable to make the payment.
But this is going to be a good litmus test for whether people will invest in this future? Will people pay for the future of web browsers? Based on OpenAI’s numbers, it seems like the answer is a resounding “yes,” but it’s going to be an interesting few months while we find out for sure.
More from Tom's Guide
- I just tested the world’s first mindful browser — it’s calmly convinced me to ditch Google Chrome
- Google will finally stop this 20-year Chrome bug from leaking your browsing history to other websites — here’s how
- More than 184 million passwords exposed in massive data breach — Apple, Google, Microsoft and more
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
