把自己綁上SpaceX戰車，Cursor的600億美元崛起之路

Original title: Inside Cursor's wild rise

Original author: Shubhangi Goel and Charles Rollet, Business Insider

Original translation: Peggy, BlockBeats

Editor's note: This article tells the story of Cursor CEO Michael Truell and the rapid rise of this AI programming unicorn.

In 2019, Truell, an 18-year-old MIT student, completed a programming test in less than 10 minutes that was originally expected to take an hour. A few years later, he co-founded Anysphere with several MIT classmates and launched Cursor, aiming to redefine how developers write code. By the end of 2025, Cursor was used by millions of developers, with revenue growing tenfold in less than a year, surpassing $10 billion.

But Cursor's story is not just a Silicon Valley narrative of a "genius programmer's entrepreneurial success." The more noteworthy part of the article lies in its revelation of the structural predicament facing AI application companies: when a company is built on top of frontier models, it can grow rapidly by leveraging those models' capabilities, but it can also be quickly squeezed out when the model provider enters the arena itself. This is precisely the relationship between Cursor and Anthropic. Cursor was once highly dependent on Anthropic's models, but after Anthropic launched Claude Code, the two transformed from partners into potential competitors, prompting Cursor to begin developing its own proprietary model, Composer.

At the same time, Cursor's high growth has been accompanied by controversy. The article mentions that Cursor's hiring process is extremely rigorous, with candidates asked to undergo multi-day or even multi-week unpaid "work trials"; internally, there has long been concern about over-reliance on a single AI model supplier. These details make Cursor's success appear more complex: it is both one of the most representative application-layer companies in the AI programming wave and a startup seeking balance between rapid expansion, an intense culture, and model dependency.

What truly pushed the story into a new phase was Truell's partnership with Elon Musk's SpaceX. To support its proprietary model, Cursor needed expensive and scarce computing power, while SpaceX/xAI needed to enhance Grok's programming capabilities. On the surface, the cooperation is a complement of computing power with data and model capabilities, but behind it lies a potential $60 billion acquisition arrangement. If the deal ultimately goes through, Cursor could become a key programming infrastructure within Musk's AI ecosystem; if it remains independent, it must prove that an AI application company can grow into a true generation-defining company in the interstices of frontier model giants.

The core question of this article is: Will Cursor become the gateway for the next generation of software companies, or just a piece of the puzzle in the AI giants' computing power war?

The following is the original translation:

Michael Truell: From Genius Programmer to Cursor CEO

In 2019, 18-year-old MIT student Michael Truell sat in the cafe of the Computer History Museum, staring at a programming test question in front of him. It was supposed to take about an hour to complete, but he finished it in less than 10 minutes.

"He completely crushed the question," recalled tech investor Ali Partovi. Partovi runs a program that identifies the world's best programmers at the undergraduate level. With plenty of time remaining, Partovi asked Truell to come up with a programming question for him instead. Partovi, himself a programmer and co-founder of Code.org, took much longer to finish. When he was done, his paper was a mess; in contrast, the teenager's lines of code were neat and clear.

Now, at 25, Truell is the CEO of Cursor. This AI programming startup has reached a potential $60 billion acquisition agreement with Elon Musk's SpaceX. The slender, frizzy-haired red-haired young man is seen by colleagues as quiet and friendly. Unlike some young founders who relish flaunting their latest revenue figures or workout results, he prefers long, almost ascetic sessions immersed in writing code. Inside Cursor, it is well known that he did not pay himself a salary for the first few years of the company's existence.

However, beneath his humble exterior, Truell has long harbored ambitions as grand as anyone in Silicon Valley. He has told employees he wants Cursor to become a "generational company." As a teenager, he developed a popular programming game themed around conquering the universe; right after graduating from MIT and starting his company, he and several college classmates challenged Microsoft in the code editor space and ultimately won. At Cursor, he fosters an intensely demanding work culture: to find the perfect fit, the company puts candidates through complex and unpaid "work trials," sometimes lasting several weeks.

Becoming one of the fastest-growing startups in the tech industry is not easy. Cursor has constantly had to navigate a delicate and tense relationship with Anthropic. Anthropic was once Cursor's primary AI model supplier, until the frontier AI lab began releasing its own highly popular programming tool. After Claude posed an existential threat to the company, Truell declared a state of emergency. Subsequently, he has increasingly tied Cursor's fate to Musk's newly public SpaceX, which is desperately trying to win the AI race and controls tens of billions of dollars' worth of computing resources.

Cursor declined to comment for this article. Anthropic and SpaceX also did not respond to requests for comment.

Truell now faces his biggest test yet: will the partnership with Musk ultimately succeed? Regardless of the outcome, the Cursor CEO is already plotting to ensure his company secures a place in the history of computing.

Growing up in New York with journalist parents, Truell was a naturally gifted programmer from a young age and began promoting programming early on. At 15, while a student at the elite private school Horace Mann, he participated in developing a programming game called Halite. The game taught programming basics by having players conquer territory on a grid. The project attracted thousands of users, mostly high school and college students who had never coded before, and earned him a $10,000 prize from a top mathematical association.

At MIT, he double-majored in computer science and mathematics and began conceptualizing startup ideas. Claire Shorall helped run a startup boot camp Truell attended during his undergraduate studies. She said she was impressed by Truell's curiosity and humility. At the time, he needed to make cold calls to doctors across the US to validate an early startup idea. Truell asked Shorall to sit next to him and help critique his communication skills as they huddled around a landline phone. That project, intended to be a competitor to ZocDoc, ultimately didn't succeed, but Shorall could already see that Truell possessed more than just raw programming ability.

"I gave him some advice—but it was obvious he already had that capability," she said.

After graduating in 2022, Truell co-founded Anysphere with MIT classmates Sualeh Asif, Arvid Lunnemark, and Aman Sanger. Initially a code editing platform, they achieved $1 million in recurring revenue in under 12 months by creating a better alternative to Microsoft's open-source code editor, VS Code.

"Our mission over the next few years is to increase programming speed by an order of magnitude while making it more fun and creative," Truell told TechCrunch at the time.

The Controversy Behind High-Speed Growth: Unpaid Trials, Extreme Hiring, and Model Dependency

To fulfill this mission, Cursor officially launched in March 2023 and grew rapidly. It quickly gained traction among developers and enterprises eager to significantly boost their productivity. In 2024, Cursor disclosed having over 40,000 customers and set an ambitious goal: to build a "magical" tool that could one day write all the world's software.

"Something beautiful is happening to code," the company wrote in a blog post at the time.

By the end of 2025, Cursor was adopted by millions of developers. The company announced that its revenue had grown tenfold in less than a year, exceeding $1 billion.

Cursor's growth was incredibly intense, and this intensity was reflected in its hiring process. Four former employees said Truell was deeply involved in hiring. He often scouted top engineers on GitHub and X, then invited candidates to Cursor's sprawling, campus-like headquarters in San Francisco for multi-day "work trials."

During the trial, candidates did almost everything regular employees would do: have lunch with the team, sit at a desk using a company computer, and complete projects based on a frozen version of Cursor's codebase.

"It really gives us a lot of signal on a candidate's raw technical ability to succeed in our environment," Truell said on a podcast in November of last year.

However, critics have pointed out that these work trials are unpaid. On Reddit, one person who claimed to have interviewed with Cursor denounced the process as "exploitative and unethical."

A former employee recalled receiving an email late at night, asking them to report to Cursor's office the next morning at 9 AM to complete a series of programming projects. In another instance, the former employee claimed Cursor put a management-level candidate through a month-long work trial. During this time, the person met nearly every team member, but the company ultimately decided not to hire them.

"After a month, their attitude was: 'We can probably find someone even better than this candidate,'" the former employee said, adding that this indicated both Cursor's extremely high standards for newcomers and the effectiveness of the screening mechanism.

Despite its phenomenal growth, Cursor's executives were long concerned that the company had become overly attached and dependent on a single AI supplier. Employees often used one word to describe the relationship between Cursor and Anthropic: strange.

The two companies were highly interdependent. Cursor relied heavily on Anthropic's AI model to power its programming tool. At the same time, Anthropic benefited enormously from Cursor's explosive growth. According to an employee familiar with the numbers, at one early stage, Cursor accounted for roughly 40% to 50% of Anthropic's revenue.

"Both sides kind of realized they needed each other. We brought in a lot of revenue for Anthropic," another employee said. "But at the same time, Anthropic had its own competing product."

Before launching its blockbuster code editor, Claude Code, Anthropic executives privately assured Cursor's management that the product was more of a research project than a major commercial push, according to a person familiar with the matter. Communications had occurred between the two parties. But Claude Code quickly went viral among developers. By February 2026, its annualized revenue had grown to $2.5 billion, about $500 million more than Cursor's annualized revenue at the time. This figure was first reported by Bloomberg. Developers also started posting that they were canceling Cursor subscriptions in favor of Claude Code.

Even before this, Cursor executives' concerns about the company's reliance on Anthropic were already high. One reason was that Anthropic had previously cut off service to competitor AI programming startup Windsurf during its acquisition talks with OpenAI.

On January 5th, Truell called what one employee described as an "emergency meeting" for the entire company, announcing that Cursor needed to build its own AI model. Two employees said the message was very clear: We must ensure we are not left behind. The company would cancel all unnecessary meetings, and people might be temporarily reassigned to work with different teams that week. We need to stay flexible and adapt quickly.

Following the meeting, Cursor began a lengthy pricing analysis comparing Claude Code and OpenAI's Codex, and also held meetings to reassure its largest customers. Executives concluded that Cursor must double down on its proprietary model to reduce dependency on frontier model labs and gain more control over pricing.

Although Cursor declined to comment for this article, Truell described the relationship with Anthropic as a "deep partnership" in a recent interview, stating they are "very grateful for it."

Cursor's Biggest Gamble: Breaking Away from Anthropic, Tying Up with Musk

Subsequently, Cursor launched Composer, its own set of proprietary programming models. Built upon an open-source model from Chinese AI lab Moonshot, Composer is already gaining traction among developers. Cursor claimed that over 85% of the Composer 2.5 model, released in May, came from Cursor's own work—meaning the underlying Moonshot model constitutes only a small part of the final product.

"Composer has received extremely positive feedback," said Cursor engineer Lucas Garza, attributing this mainly to its low price and high speed, especially against the backdrop of rising AI costs and pressure on tech companies' engineering budgets.

Cursor's latest tools are also generating new excitement. On a hot afternoon in June, Cafe Cursor, a pop-up shop run by Cursor in San Francisco's North Beach tourist district, was arguably the busiest cafe on the block. The pop-up offered free lattes and $50 credits to enthusiastic entrepreneurs, many of whom praised Cursor for boosting their productivity.

This month, tech workers lounged at Cafe Cursor, a pop-up coffee shop under the Cursor brand. Charles Rollet/Business Insider

Aneesh Dharani, founder of an AI flashcard startup, said that despite having no software engineering background, Cursor helped him actually build his product. Another founder, Devon Lim, said he replaced an outsourced engineer with Cursor after the engineer suddenly went "off the grid" and stopped working for his sales startup.

But building and running a top-tier AI model is prohibitively expensive, and Cursor lacks sufficient chips to do it entirely independently. So, this spring, Truell and his company found another founder with "interstellar-level ambition" to fill the gap: Elon Musk.

On April 21st, Truell announced a new collaboration on X in his typically concise style.

"Excited to work with the SpaceX team to scale Composer. A big step in our journey to build the best place for AI programming," he wrote.

On the surface, the deal benefits both sides. Cursor gains access to SpaceX's massive computing power resources, including Colossus—a supercomputer powered by hundreds of thousands of top-tier Nvidia AI chips. Meanwhile, SpaceX's Grok gets a boost in the AI programming competition. A contractor for xAI previously told Business Insider that Grok was not the "best at programming."

What Truell didn't mention in that X post was a much bigger development: he had already agreed that SpaceX might acquire Cursor for $60 billion later this year.

This news surprised many Cursor employees, as Truell had consistently talked about building Cursor for the long term. One former employee said that whenever an acquisition was mentioned, Truell would say, "This is a huge risk we are taking, or a huge bet."

The structure of the deal is also unusual. According to SpaceX's S-1 filing last month, if either party decides not to proceed with the transaction, SpaceX will pay Cursor a $1.5 billion termination fee and provide an additional $8.5 billion in free computing power.

Ali Partovi, one of Cursor's earliest investors, is not privy to the deal's internal details. He said that while many founders claim they would never sell their company, in reality, they fall on a spectrum. Partovi believes Truell leans more towards the end that favors staying independent.

"His ambition, confidence, and drive would push him more towards independence," Partovi said.

For now, Cursor remains independent and continues to grow rapidly. According to Forbes, its revenue doubled in three months, reaching $4 billion.

Some early progress has been made. Musk posted on X that a recent version of Grok had improved significantly after being trained on a "huge amount" of Cursor data. Both Grok and Composer are climbing the ranks in closely watched AI model leaderboards—i.e., benchmarks—though neither has reached the top yet.

For Musk, the goal is clear: his AI will become "very strong" regardless.

"Whether it will become the strongest remains to be seen, but I will never give up," he wrote on X. "Never."

For Cursor, the ultimate goal is less clear, as the structure of the deal with SpaceX remains quite open-ended.

Truell stated in a recent interview that Cursor now has 700 employees and serves 60% of Fortune 500 companies. He added that the company can now be compared to many of the largest publicly traded software companies in the world.

"It is indeed a bit crazy," he said. "And we are also very aware of how special this is—how unprecedented it is from a historical perspective."

Original link