Cursor announces major update as AI coding agent battle heats up

Cursor announced Tuesday that it is making updates to its AI coding agents as it tries to fend off increasingly stiff competition from rivals like Anthropic, OpenAI and others. Microsoft.

The AI ​​startup needs to have cutting-edge technology to gain new users and market share, and the company told CNBC that its updated reps take its capabilities to the next level. The startup’s valuation rose to $29.3 billion, and in November it was announced that its annual revenue exceeded $1 billion. Although Cursor entered the AI ​​coding market relatively early, other players were quick to launch competing offerings.

AI agents are tools that can complete tasks on behalf of a user and have skyrocketed in popularity over the past year as their model capabilities have improved. Some of the early adopters were software developers who used agents from companies like Cursor to create, edit and review code.

Cursor’s updated agents will be able to test their own changes and record their work through videos, logs and screenshots, the company said. Users can trigger agents from the web, Cursor’s desktop app, mobile device, Slack messaging platform, or Microsoft’s GitHub.

Agents can also run in parallel with full development environments on their own virtual machines, which are cloud-based computers that behave just like a physical computer. This means reps won’t have to compete locally for resources on a developer’s laptop or waste time recruiting, Cursor said.

Asynchronous agents “Instead of one to three things you’re doing at the same time running at the same time, you can have 10 or 20 of them running,” Alexi Robbins, Cursor’s co-head of engineering, told CNBC in an interview. “You can get a really high yield with this.”

As of February, Anthropic’s Claude Code run rate revenue had surpassed $2.5 billion, and OpenAI Codex’s weekly active users had surpassed 1.5 million. Microsoft CEO Satya Nadella said in October that the GitHub Copilot tool has more than 26 million users.

Cursor was founded in 2022 and garnered a cult following after launching the first version of its AI coding product the following year.

Developers can delegate more complex tasks to Cursor’s updated agents, which can test and iterate on these features until completion. This will allow developers to spend less time editing files and parsing code and more time focusing on questions of taste and judgment, Cursor said.

“We’re thinking of this less as a new feature and more as ‘This is what working with delegates will look like,'” said Jonas Nelle, Cursor’s other vice president of engineering for asynchronous delegates. “They don’t just write software, they don’t write code, they kind of become full software developers.”

Cursor is testing agents internally, and Nelle said it’s already been a “huge transformation” for the company.

Cursor said that about 35% of Cursor’s pull requests, where a developer suggests a change to the codebase, are now created by agents running on their own virtual machines.

“You can do so much more by working with these agents individually,” Nelle said. “This is another step change in that progression.”

WRISTWATCH: Cursor CEO Michael Truell: We see success not only with demand, but also in our age of artificial intelligence