
Browser from Opera for handling tasks...
When sources are still scattered across tabs, finished research usually has to be moved into another tool. Neon reads pages, keeps context, and compares inputs inside the same browser that already holds the tabs, sources, and notes. Assistants split the work: some draft an agenda, some browse, and some do deeper research or documents, without rebuilding the flow each time. An agent you already use can connect through MCP or the command line and work on the page, rather than in a separate window outside the browser. Ad blocking and a VPN follow Opera's built-in approach, and credentials, passwords, and cards stay on the device. Account data, chats, and browsing history are not used for training or ads. It fits people who want to deliver a draft while they browse, and users who want an existing agent attached to page context.
Opera Neon is a browser from Opera for handling tasks, reading web pages, and managing tabs while you browse. It keeps context, compares sources, and turns the result into something you can send or keep editing. Different assistants in the browser can draft an agenda, browse, do deeper research, or generate a document. Agents you already use can connect through MCP or the CLI and bring the current page context with them. Ad blocking and a VPN are built in. Agentic features turn on after a subscription. Login credentials and card details stay on the device.
Difficulty: Intermediate
Organize research in the browser
It reads page content, keeps the context of tabs and notes, compares inputs, and turns the result into material you can send or keep editing.
Call different assistants by task
Several assistants can live in the browser and handle an agenda, web browsing, deeper research, or document generation without rebuilding the flow each time.
Connect an existing agent
An agent you already use can connect through MCP or the CLI, bring browser context, and run a workflow on the page itself.
Built-in blocking and local credentials
Ad blocking and a VPN are built in. Login credentials, passwords, card details, and cookies stay on the device and are not used for model training or ads.
Turn several pages into a draft
A researcher compares sources and notes in tabs that are already open, and has the browser turn the raw material into a draft that can be sent.
Let an existing agent work on a page
A developer connects the agent they already use to Neon through MCP or the CLI so it can act with the current page context.
Block ads and use the VPN while browsing
Someone browsing day to day uses the built-in ad blocker and the VPN that does not need an extra extension, without subscribing to agent features first.
Not automatically. The start steps on the site are download, install, and log in. If you already use an agent, connect it through MCP or the command line. Agentic features and the major language models named on the page turn on only after you subscribe. The built-in VPN is separate: the page says it does not need a subscription, payment, login, or an extra extension. The ad blocker is built in as well, and a site can be allow-listed.
The site says account data, chats, browsing history, and similar data are not used for model training, personalization, or ads. Login credentials, passwords, card details, and cookies stay on the device. At the same time, it uses your location, preferred online services, and the pages you are viewing so it can help with the task. Both statements are on the page, so it is worth separating what stays local from what is used to understand the current task.
The page stresses that it is a browser first and includes a set of everyday Opera abilities: smarter tab management, snapshots, messaging and services in the sidebar, modular browsing, a built-in ad blocker, and a VPN. The agentic web is added on top of those. Opening pages, blocking ads, and switching tabs remain the base even when you use it for tasks.
Real reviews and feedback from users