Cursor launches AI agent experience to compete with Claude Code and OpenAI Codex

Cursor Launches a New AI Agent Experience to Take on Claude Code and Codex

Cursor announced Thursday the launch of Cursor 3, a new product interface that allows users to spin up AI coding agents to complete tasks on their behalf. The product, which was developed under the code name Glass, is Cursor's response to agentic coding tools like Anthropic's Claude Code and OpenAI's Codex, which have taken off with millions of developers in recent months. "In the last few months, our profession has completely changed," said Jonas Nelle, one of Cursor's heads of engineering, in an interview with WIRED. "A lot of the product that got Cursor here is not as important going forward anymore." Cursor increasingly finds itself in competition with leading AI labs for developers and enterprise customers. The company pioneered one of the first and most popular ways for developers to code with AI models from OpenAI, Anthropic, and Google -- making Cursor one of these companies' biggest AI customers. But in the last 18 months, OpenAI and Anthropic have launched agentic coding products of their own, and started offering them through highly subsidized subscriptions that have put pressure on Cursor's business. While Cursor's core product lets developers code in an integrated development environment (IDE) and tap an AI model for help, new products like Claude Code and Codex center around allowing developers to off-load entire tasks to an AI agent -- sometimes spinning up multiple agents at the same time. Cursor 3 is the startup's version of an "agent-first" coding product. According to Nelle, the product is optimized for a world where developers spend their days "conversing with different agents, checking in on them, and seeing the work that they did," rather than writing code themselves. Cursor is launching its new agentic coding interface inside its existing desktop app, where it will live alongside the IDE. At the center of a new window in Cursor, there's a text box where users can type, in natural language, a task they'd like an AI agent to complete -- it looks more like a chatbot than a coding environment. Press enter, the AI agent sets to work without requiring the developer to write a single line of code. In a sidebar on the left, developers can view and manage all of the AI agents they have running in Cursor. What's unique about Cursor 3, compared to desktop apps for Claude Code and Codex, is that it integrates an agent-first product with Cursor's AI-powered development environment. In a demo, Cursor's other cohead of engineering for Cursor 3, Alexi Robbins, showed WIRED how users can prompt an agent in the cloud to spin up a feature, and then review the code it generated locally on their computer. Nelle and Robbins argue it doesn't matter which interface developers are spending their time in -- they just want people using Cursor. I visited Cursor's office in San Francisco's North Beach neighborhood last week. The startup is reportedly raising fresh capital at a $50 billion valuation -- nearly double what it was valued in a funding round last fall -- and has expanded into an old movie theater. Cursor employees used to toss their shoes in a pile by the door upon entry, but now there's a row of large shoe racks, signaling one way in which the company is growing up. Yet Cursor still feels like a startup. Employees tell me that's part of the appeal of working there; the company can ship quickly and doesn't feel too corporate. But as it finds itself racing to catch up to Anthropic and OpenAI in the agentic coding race, that scrappiness may not be enough. This battle -- the one to create the best AI coding agent -- may be Cursor's most capital-intensive chapter yet.

Cursor's New Tool Lets Users Delegate to a Team of Coding Agents

With Anthropic's Claude Code and OpenAI's Codex gaining traction as coding assistants, Cursor has been feeling the pressure to ramp up its own AI-powered coding offerings. Enter Cursor 3, the company's new environment for creating and managing multiple AI agents working on your behalf. Per the company, Cursor 3 is a "unified workspace for building software with agents." Less of a model overhaul and more of a new interface, the new release seeks to turn the user into something more like a manager who can dictate orders to a team of AI agents. According to the company, Cursor 3 can be used to manage both local and cloud-based agents, and allows users to work across multiple repositories on a project. The shift that Cursor 3 seems to present is more of an embrace of vibe coding at a time where other leading AI firms are increasingly encroaching on the territory once dominated by Cursor. The company's core product, which is still available through the new environment, lets developers ask AI assistants for help in an integrated development environment. This interface lets those developers take a step back from the coding tasks and do more delegating to multiple agents while providing a 30,000-foot view of the work. The launch of Cursor 3 comes at time when the company really could use a win. Claude Code has reportedly captured up to 54% of the AI coding market, according to data from Menlo Ventures, and OpenAI's recently released Codex 5.3 set new highs in a number of benchmarking testsâ€"and the company is offering unlimited access to the tool in order to pull in more users. Cursor still remains a popular tool among product and engineering teams, but its reign as the only game in town is long gone. The company could also use a bit of reputational repair after the bumpy launch of Composer 2, the company's underlying coding model that dropped last month. The model was supposed to be a marker of Cursor separating itself from some of its competitors, but instead managed to link the company even closer to a third-party once it was discovered that Composer 2 was largely just a licensed version of the open-source Kimi 2.5 model made by Moonshot AI. In general, there's nothing wrong with licensing another model this way, but the fact that Cursor didn't disclose it up front and seemed to try to bury it has made some users a bit wary of the company's tactics.

Cursor refreshes its vibe coding platform with focus on AI agents - SiliconANGLE

Startup Cursor today debuted a new version of its popular artificial intelligence coding platform. The release includes features that will make it easier for developers to automate programming tasks using AI agents. It also introduces improvements in other areas. Cursor, officially Anysphere Inc., has raised more than $3 billion in funding from Nvidia Corp., Google LLC and other prominent backers. It provides a code editor that uses AI to automate tasks such as building new application features and debugging existing ones. The tool is powered by not only off-the-shelf large language models such as Claude but also algorithms that Cursor has developed in-house. Cursor 3, the new platform release that debuted today, adds a chatbot interface. It enables developers to enter a natural language description of the feature they wish to build and specify the LLM that should carry out the task. Cursor 3 generates the requested code along with a demo video that shows how it works. The new chatbot interface uses multiple AI agents to complete user-specified tasks. Some of those agents run in the cloud, while others are installed on the user's local machines. Developers can centrally manage both agent varieties through a newly added sidebar. Cursor's cloud-based agents have access to more hardware resources than ones that run on workstations. As a result, a large number of them can run in parallel to speed up time-consuming tasks. Desktop agents are slower, but they enable developers to locally open the code being generated, edit it manually and run tests. Cursor 3 enables users to switch between the two modes. A developer could have a set of cloud-based agents generate a piece of code and then send it to a desktop-based agent for local editing. Cursor says that Composer 2, an internally-developed LLM it debuted last month, is particularly well suited for such tasks. The model is more cost-efficient than several of the other LLMs supported by the platform. When developers are using Cursor to edit a user interface, they can activate a newly added Design Mode to speed up the workflow. The feature makes it possible to select interface elements and add natural language text that describes how they should be changed. Cursor's AI agents automatically implement the requested modifications. The platform displays a step-by-step overview of each task that it completes. Cursor provides a natural language explanation of every sub-step, highlights any errors that may have cropped up and provides screenshots of its work. Developers can adjust the workflow by typing in feedback.

Cursor unveils new AI agent to take on Claude Code and OpenAI Codex: What it can do

Cursor now competes with tools like OpenAI Codex and Claude Code, while also building its own models like Composer 2. Cursor is all geared up to redefine how developers interact with artificial intelligence, as the company has now launched the latest version of its code editor, Cursor 3. In this latest version, the company has redesigned the interface to incorporate AI agents that can perform coding jobs on their own. Earlier, the company was famous for its AI-powered assistance and autocomplete features. However, the company has now shifted its approach toward agentic coding, which allows developers to perform operations autonomously rather than relying on manual line-by-line assistance, thus streamlining the workflow for developers across the globe. The new Cursor 3 interface allows users to describe tasks in plain language through a central Agents Window. Once a request is submitted, an AI agent begins working without requiring manual coding; as a result, managing parallel tasks becomes easier as developers can now track multiple agents across different repositories at once. Unlike standalone tools, Cursor combines this agent system with its existing development environment, letting users review and refine generated code locally with a high degree of control. Also read: Apple 50th anniversary sale: iPhone 17 Pro Max, iPhone 17 Pro, iPhone gets big price cut This has put Cursor in direct competition with companies such as OpenAI and Anthropic, which have recently released similar agent-based tools. For example, OpenAI's Codex App and Claude Code have been quite popular, especially because they offer competitive efficiency and reasoning. These tools have started to gain popularity among developers for making significant changes in their architecture. However, Cursor has been under pressure because of a significant change in their pricing strategy, which took place in 2025. The tech firm has moved from a flat subscription-based model to a usage-based credit system. Although this has been positive for Cursor's sustainability, there has been a degree of dissatisfaction among users because of this change. Also read: OpenAI admits it is skipping opportunities due to lack of compute In order for Cursor to remain at the top, they have started creating their own proprietary models to remain at par with their competitors. For example, Composer 2 is a testament to this, as this model has been optimized for agentic tasks, balancing speed, cost, and performance within the editor. Reports suggest the firm plans to invest even more in creating in-house models to reduce reliance on external providers, even though this remains a costly affair.