Following its official announcement in May, Opera will today start giving access to Neon, its subscription-based AI browser with agentic capabilities. Here are some of its most interesting features. With Neon, Opera joins the incipient, but increasingly competitive, AI-based browser market. And in an attempt to stand out, the company has developed a few interesting features that will help users make the most of Neon’s agentic capabilities. Tasks First up is Tasks. It’s designed to understand the context of the tabs you have open, whether that’s a document, a webpage, or a search, and use that information to gather details and perform the action you’re trying to complete. Here’s how Opera describes the feature: “Tasks are self-contained workspaces that understand context and make it possible to use the AI to analyze, compare, and act across multiple sources at once. You can think of it as Opera Neon creating a mini-browser for each of your tasks, where the AI understands what you’re doing and helps you within this context—without accessing information from everything else in your browser.” Here is an example of the feature in action, where the user requests Neon to group and compare notes between Notion, Google Docs, and Gmail: Cards This is perhaps the most promising idea of Neon. With Cards, users can save the prompts they use more frequently, rather than typing them in from scratch for repetitive tasks. Users can also mix, match, and daisy-chain Cards for more complex tasks. Here’s Opera: “It’s like having a deck of your favorite Al behaviors ready to use when you need them. Comparing products across tabs? Add the pull-details + comparison-table cards to your prompt. Or if you’re taking meeting notes, combine the key-decisions + action-items + follow-ups cards and Opera Neon will capture what matters in the right format.” Another interesting aspect of this feature is the Cards store, which lets users browse and save community-uploaded cards. If you’re a developer and you’re thinking “functions,” that’s pretty much it. But Cards add a visual element to that concept, making it much more approachable to casual and pro users alike. Neon Do This feature does what is becoming known as agentic browsing, but with an interesting twist: rather than running a browser session in the cloud, Neon Do works within your actual browser session. This means that Neon Do can leverage the fact that you’re probably already logged into the services and platforms you use the most, so it can actually navigate on your behalf, gather data from your actual context, and work on the task you assign it. Or, as Opera puts it: “When you activate Neon Do within a Task, it starts operating inside your browser session, where you’re already logged in. There is no need to share passwords with cloud services or for repeated authentication flows. Neon Do navigates the real web on your behalf – checking multiple sites, comparing information, filling out forms, gathering data from pages in your Task. Some actions may require user interaction – that’s when Neon will pause and wait for you to act.” When a specific action requires user interaction, such as logging into a website, Neon Do pauses its workflow and requests that the user take over. Users can also interact with the agent mid-task, or take over at any point. Aside from these three features, there is also Make, which can build small apps based on the requests and the needs of the user, and Chat, which lets the user interact with Neon based on the context of the webpage they are viewing. Opera says that Neon was built with a privacy-first approach. For this reason, logins and payment information stay on-device, and perhaps more importantly, nothing is used for training. Opera Neon requires a $19.99 monthly subscription, and the first users will gain access today. You can visit Opera Neon’s website to join the waitlist. Are you interested in agentic browsing? Let us know in the comments. Accessory deals on Amazon