Pulse News

FridayMarch 14, 2025

The Math Behind a $10 Billion Valuation For Cursor

View Original Article →Published: 3/7/2025

**The Math Behind a $10 Billion Valuation For Cursor**

By Natasha Mascarenhas and Stephanie Palazzolo

Mar 7, 2025, 1:51pm PST

Now that the word is out about the maker of Cursor, an artificial intelligence-powered coding assistant, talking to investors about a new round that would value it at $10 billion, it's important to explain why a deal may happen at that price.

The new valuation—up four times from a deal struck just three months ago—would actually be in line with other recent investments involving high-profile AI startups, as a multiple of their revenue. That's because the San Francisco-based Anysphere tripled its annual recurring revenue to more than $150 million in the past four months, according to a person with direct knowledge of the company's finances. That figure implies it is generating about $12.5 million in monthly subscription revenue. (As you can see from a chart in the middle of this piece, a multiple of 66 times forward revenue isn't crazy.)

A deal would further elevate the profile of Anysphere in the super-hot field of AI for software developers and give it ammunition in its fight against OpenAI and Anthropic, which sell competing products. Anysphere doesn't develop its own generative AI and instead depends on models from Anthropic and OpenAI to power Cursor. But venture capitalists have been willing to back app developers despite their reliance on AI model makers, largely due to their fast revenue growth, which is outpacing an earlier generation of software startups.

Cursor's core service helps software developers at companies such as card startup Ramp and AI search engine Perplexity write code with autocomplete and other editing services. It's one of three classes of coding assistants that have sprung up since the launch of OpenAI's ChatGPT—an app that is also widely used by software developers for AI-powered coding.

Developers have praised Cursor for its ease-of-use, speed, and ability to reference and understand a user's codebase better than rival products. With free, $20 per month, and $40 per month tiers, it's also cheaper than rival products like Devin, which costs $500 per month and aims to tackle more complex coding tasks.

Anysphere is just over three years old. In December, it raised $100 million in funding at a $2.5 billion valuation from investors including Thrive Capital—a big backer of OpenAI—Andreessen Horowitz, and Benchmark. Other investors in the company include OpenAI's startup fund and Neo, a startup accelerator.

Representatives for Anysphere didn't respond to requests for comment. Bloomberg earlier reported on the deal talks, which we've also heard would value the company at around $10 billion from people who've spoken to investors in the company.

While Cursor initially used OpenAI's models to power its coding editor, it changed its default model to an Anthropic one in July. A few months later, OpenAI released its own code editing tools within ChatGPT through a feature called Canvas. Then last week, Anthropic launched its own coding editor, Claude Code, as it seeks to generate more subscription revenue from such services. OpenAI is also developing an even more ambitious coding assistant product that aims to replicate the work of a senior software engineer, The Information has reported.

VCs have latched onto other coding assistant startups, hoping the exploding market is big enough for all of them. Last month, Kleiner Perkins led a round into Codeium at a nearly $3 billion valuation. Similar to Cursor, the Codeium deal comes just months after the coding assistant startup raised $150 million at a $1.25 billion valuation.

And in recent weeks, investors have been talking to executives at Poolside, which is developing a coding assistant app and coding assistant models, about investing more money in the company despite it generating less than $10 million in revenue last year. The company could target a $5 billion valuation in an eventual fundraising.

**Here's what else is going on...**

Did Musk Get a 'Win' in This Week's OpenAI Ruling?

Earlier this week, a U.S. district court judge denied Elon Musk's request for a preliminary injunction to stop OpenAI's transition to a for-profit model from its current position as a nonprofit charity. That might not sound like a great result for Musk, but his legal team seems excited that the judge in the case didn't shoot down his legal standing to bring a claim of breach of charitable trust against OpenAI. Musk claims that OpenAI executives misled him into believing that the company would remain a nonprofit focused on developing AI for the benefit of humanity, but later transitioned toward profit-driven goals in partnership with Microsoft.

Judge Yvonne Gonzalez Rogers said in her ruling that she would allow an expedited trial over that claim. The main question, she said, is whether Musk's donations legally created enforceable restrictions on OpenAI's business model, especially since there aren't any formal contracts or trust documents. If Musk can prove that other evidence, such as the emails he exchanged with OpenAI executives, constitute the formation of a charitable trust, the breach claim would tip in his favor, she said. Also, expediting the trial to fall of this year, as the judge decided, means the case won't drag on for years as OpenAI was hoping for.

**Authors' Information**

Natasha Mascarenhas is a reporter at The Information, based in San Francisco, who covers venture capital and startups. She can be reached at natasha@theinformation.com, or on Signal at +1 925 271 0912. She is on Twitter at @nmasc_

Stephanie Palazzolo is a reporter at The Information covering artificial intelligence. She previously worked at Business Insider covering AI and at Morgan Stanley as an investment banker. Based in New York, she can be reached at stephanie@theinformation.com or on Twitter at @steph_palazzolo.