In late May 2025, Opera unveiled Neon—its latest attempt to redefine how we interact with the web—this time through an “agentic” AI browser that promises to handle tasks for you, even while you’re asleep. Dubbed an “agentic browser,” Opera Neon is built around AI agents designed to understand user intent, leverage contextual awareness, and perform complex browsing tasks on behalf of users. Opera envisions Neon not merely as a browsing tool but as a collaborative partner capable of researching, building, designing, and coding anything you need, without you having to lift a finger.

Oddly enough, Opera first introduced a concept browser called Neon back in 2017, aiming to showcase futuristic design elements like picture-in-picture windows and intelligent tab management. Despite turning heads, that version never gained traction and was quietly shelved. Fast forward to 2025, and Opera has resurrected the Neon name with entirely new ambitions. This iteration pivots away from design experiments toward AI-driven “agentic browsing,” reflecting how far both Opera and AI technology have come in the intervening years.

At its core, an agentic browser is one that doesn’t just respond to clicks and keystrokes but proactively interprets and executes user instructions. Opera Neon’s AI engine is described as “capable of understanding and interpreting” what users want and then working with cloud-based AI agents to fulfill those requests—even operating in the background when the user is offline. For example, if you ask Neon to build a simple website or generate a snippet of code, the AI agents will marshal resources to draft HTML, CSS, or JavaScript files for you. If you need a custom game prototype, Neon’s agents will research libraries, pull together assets, and assemble a playable demo—all while you’re catching some Z’s.

Opera senior AI product director Henrik Lexow put it plainly in the company’s press materials: “We’re at a point where AI can fundamentally change the way we use the internet and perform all sorts of tasks in the browser. Opera Neon brings this to our users’ fingertips.” While Opera hasn’t provided a firm release date yet, it has opened a waitlist for those curious to test-drive Neon when its beta arrives.

Opera is quick to draw parallels between Neon’s AI suite and other products like Microsoft’s Copilot and OpenAI’s Operator. Indeed, Neon features a chatbot interface that lets users ask natural-language questions, search the web, or extract “contextual information” directly from open pages. Moreover, it incorporates Browser Operator—an AI agent Opera previewed in March 2025 that automates routine web tasks such as shopping, filling out forms, and booking events or accommodations. Unlike cloud-based solutions, Browser Operator runs natively on the user’s device, utilizing the browser’s DOM tree to interact with page elements rather than relying on screenshots or virtual machine sessions. This approach not only boosts speed and reliability but also strengthens privacy by keeping user data local until explicitly needed.

Still, Opera seems keenly aware that simply grafting AI onto a browser isn’t enough; it’s the “agentic” element that sets Neon apart. Where Copilot and Operator serve primarily as passive assistants—waiting for your prompt to summarize a document or draft an email—Neon’s agents can autonomously pursue tasks from start to finish.

Opera Neon is organized around three core modes: Chat, Do, and Make. Each mode is optimized for a different class of tasks:

• Chat: A conversational interface akin to a chatbot, where users can ask questions about the open webpage, request facts, or seek real-time web search results. For example, you could ask Neon to summarize the key insights from a lengthy research article and get a concise bullet-point list in seconds.
• Do: The true agentic engine. In this mode, Neon takes directives like “Book me a budget-friendly flight from Mumbai to New York next month” or “Research and draft a two-page market analysis report on electric scooters.” Neon’s agents scrape relevant websites, compare options, fill out booking forms up to the payment stage (pausing for your confirmation of sensitive inputs), and then present a deliverable. Opera claims that “Do” mode can even create simple video games by sourcing assets, writing code scaffolding, and packaging the game for download.
• Make: A hybrid mode tailored for content and code creation. If you need a snippet of JavaScript to implement drag-and-drop functionality or a responsive CSS grid layout for a personal blog, “Make” mode can generate and test that code locally. It also assists in composing long-form content such as blog posts, project proposals, or press releases. Essentially, “Make” is aimed at bridging the gap between ideation and execution for creative or technical tasks.

Importantly, all these modes run through cloud-based AI agents but leverage local computation when possible. For privacy-sensitive actions—like logging into your bank account or submitting payment details—Neon’s Browser Operator pauses and prompts you to manually enter credentials. Only after you confirm that step will the AI proceed, and it does so without ever sending keystrokes or screenshots to Opera’s servers. This creates a balance between autonomy and user control, a tightrope Opera hopes Neon will walk successfully in production.

Many early demos of Neon emphasize that Browser Operator runs natively within the browser rather than in a cloud-based VM. According to an Opera press release, the agent uses the browser’s Document Object Model (DOM) and layout data to understand pages’ structure, enabling it to fill forms or click buttons without needing to “see” the screen pixel by pixel. This design choice yields three significant advantages:

1. Privacy preservation: Because Neon doesn’t rely on screenshot-based UI automation, no sensitive data—like passwords or credit card numbers—ever leaves your device. When you input such information, Browser Operator pauses and hands control back to you, resuming only once you’ve completed the step.
2. Speed and efficiency: Interacting directly with the DOM allows Neon’s agents to process entire pages in bulk, avoiding the need to scroll or interpret visual elements manually. As a result, tasks that might take minutes to perform manually can be wrapped up in seconds of AI-driven work.
3. Robustness across sites: Traditional automation scripts often break when a website’s layout changes. By reading the underlying HTML structure rather than pixel data, Neon’s AI is more adaptable, maintaining functionality across site updates and designs. Opera has touted this resilience as a major selling point compared to existing browser automation tools.

For now, Opera isn’t divulging pricing details or a concrete launch date. Neon is positioned as a “premium subscription product,” which suggests that access to its advanced AI features will be locked behind a paywall. Interested users can sign up for Opera’s waitlist on Neon’s dedicated website, and early adopters will gain access to a closed beta. However, with so many AI-powered browsers in the pipeline—including Microsoft’s ongoing Copilot integrations, Google’s AI-first Chrome experiments, and new entrants like Arc’s Dia and Perplexity’s Comet—Opera will have to balance innovation against cost to justify a subscription fee.

Interestingly, Neon is the fifth browser in Opera’s 2025 lineup. In February, Opera released Air—a mindfulness-focused browser that offers built-in breathing exercises, break reminders, and calming soundscapes alongside standard features like a free VPN and ad blocker. The rapid-fire rollout of Air, followed by Neon, underscores Opera’s ambition to carve out niche markets rather than directly competing with established names like Chrome or Firefox on feature parity alone.

Skeptics note that we’ve seen ambitious “AI browser” claims before. In 2017, Neon’s predecessor generated excitement with its avant-garde design but ultimately sputtered due to a lack of clear use cases. The 2025 Neon faces the inverse challenge: its use cases are compelling, but its execution must meet high expectations. Will Neon be robust enough to navigate Amazon’s constantly shifting checkout flows? Can it seamlessly book complex multi-city flights without tripping over captcha prompts or two-factor authentication? Only a blind beta test will tell.

Moreover, enterprises and privacy-conscious users will scrutinize Neon’s promise of local execution. Although Opera insists that no private data ever leaves the device, the act of letting an AI agent “browse and act” on your behalf could be unsettling. Opera will need to build trust by making Neon’s internal logs and actions transparent so users can audit exactly what the AI is doing at each step.

Opera Neon’s ambition signals a broader shift in how we conceive of the browser. For over three decades, browsers have been passive gateways to the web, rendering HTML, CSS, and JavaScript while waiting for user input. Neon suggests a future where the browser becomes an active collaborator—an application in its own right that both interprets and executes user intent. As Opera EVP Krystian Kolondra put it during the Mobile World Congress preview, “For more than 30 years, the browser gave you access to the web, but it has never been able to get stuff done for you. Now it can.”

If Neon fulfills its promise, it could redefine online workflows. Developers might shift from writing boilerplate code to guiding Neon’s “Make” mode to generate scaffolding. Remote workers could let Neon handle scheduling meetings across time zones or booking hotels for entire departments. Even casual users might delegate vacation planning—Neon could comb through multiple airlines, compare hotel rates, and suggest itineraries, handing over final booking links for human confirmation. The time-saving potential is vast.

Yet, this agentic future raises philosophical questions: If AI can do everything we need online, what’s left for us to do? Opera seems poised to address this by keeping humans “in the loop” for sensitive steps while letting Neon tackle the mundane. As AI capabilities continue to advance, we may see more browsers—from Chrome to Safari—experiment with similar agentic features. For now, Opera Neon stakes its claim as the pioneer in this space, and eagerly watching how its AI-driven promise holds up in real-world use will be worth the wait.

By marrying AI with native browser architecture, Opera Neon aims to usher in an agentic web era—one where your browser is less a passive tool and more a proactive partner. Whether it succeeds where its 2017 namesake faltered remains to be seen, but the promise of code generation, automated booking, and multi-tasking AI agents working round the clock certainly makes Opera Neon one of 2025’s most intriguing tech experiments.