Michael
Truell

He was 11 years old, growing up in New York City, and the game he wanted to play didn't exist yet. So Michael Truell built it. That instinct — close the gap between what you want and what exists — turns out to be the one trait that scales from a Bronx bedroom to a $29.3 billion company.

By 14, Truell was at Horace Mann School in the Bronx and co-creating Halite, an online programming competition with classmate Benjamin Spector that would eventually draw more than 5,500 participants — many of them college students, some of them learning to code for the first time. The project won the 2017 ACM/CSTA Cutler-Bell Prize in High School Computing, a $10,000 award and one of the most prestigious recognitions in student computer science. He was also winning medals at the International Olympiad in Informatics, one of the most competitive programming contests in the world.

None of this is the interesting part. The interesting part is what he did next: went to MIT, interned at Google building language models for feed ranking, met Ali Partovi (the early investor behind Facebook and Airbnb) by completing a written coding test "in record time," became one of 30 annual Neo Scholars — and then, in 2022, dropped out of MIT to build something that had no name yet.

The company they incorporated was called Anysphere. The product they built was called Cursor. The premise was maximalist to the point of absurdity: don't build a plugin for an existing editor. Don't add AI to VS Code. Build a new editor from the ground up — one where AI is structural, not cosmetic.

They launched publicly in March 2023. By January 2025 — twelve months after reaching meaningful scale — Cursor had crossed $100 million in annual recurring revenue. No SaaS company in history had grown that fast. Six months later it was $500M. Then $1B. By February 2026: $2 billion ARR.

In 2022, Michael Truell, Aman Sanger, Sualeh Asif, and Arvid Lunnemark were nearing the end of their time at MIT. They were building in AI before the term "AI startup" became a cliche — before ChatGPT rewired the industry's imagination in November 2022. They had already been exploring a "copilot for mechanical engineers." They tried it for six months. It didn't feel right.

Then they looked at software development itself — the profession they were trained in, the one they loved — and asked what AI could do to it. The honest answer was: almost everything. They saw it coming. The intimidating part was that everyone else saw it too. AI coding tools seemed "too competitive" to enter. That hesitation lasted about one conversation.

"We realized we were really inherently excited about the future of coding," Truell has said. Excited enough to drop out. Excited enough to not add features to VS Code, which would have been faster, cheaper, and strategically inferior. They built their own editor.

The seed came from the OpenAI Startup Fund in October 2023 — $8 million, alongside early checks from GitHub's former CEO Nat Friedman and Dropbox co-founder Arash Ferdowsi. Andreessen Horowitz led the Series A in July 2024. Thrive Capital, Google, and Nvidia followed. The raises got bigger. The intervals got shorter. In November 2025, Accel and Coatue co-led a $2.3 billion Series D at a $29.3 billion valuation — five months after the previous round.

In April 2026, SpaceX (through Elon Musk's xAI) secured an option to acquire Cursor for $60 billion — or pay $10 billion for a deep collaborative arrangement. The MIT dropout who first coded to make games was now fielding acquisition offers larger than the GDP of most countries.

Truell's vision isn't "GitHub Copilot but better." It's categorically stranger. He talks about a future where the representation of software logic looks more like English than Python — where the editor becomes an interface between human intent and machine execution, not a place to type imperative instructions.

This shaped every product decision from the start. Cursor was built as a standalone editor rather than an extension because Truell believed all programming would eventually flow through AI models. If that's true, the editor's architecture matters at a fundamental level. You can't retrofit.

Where most AI tools measure daily active users, Cursor's team measured something they called paid power users — developers who code four or five days a week and actually pay. "If someone's willing to pay for your product, they think it's genuinely useful," Truell has noted. The metric kept the team honest about real adoption instead of flattering vanity numbers.

He's also candid about the risks. When the "vibe coding" movement — building software by delegating blindly to AI without reading the output — gained attention in 2025, Truell issued a pointed warning: "If you close your eyes and you don't look at the code, and you have AIs build things with shaky foundations as you add another floor, and another floor, and another floor — things start to kind of crumble." The thing that will separate good engineers from fragile ones isn't raw syntax fluency. It's judgment. Taste.

On competition from the same model providers his product relies on — OpenAI, Anthropic — Truell is characteristically unfazed. He compared any product from these companies to "a concept car engine without the surrounding vehicle." What Cursor does, he argued, is take the best intelligence from multiple providers and build the full car around it. The end-to-end matters.

Ali Partovi (Neo Scholars) — spotted Truell during his Google internship and committed to backing him before he had a company idea. Nat Friedman (former GitHub CEO) and Arash Ferdowsi (Dropbox co-founder) wrote early angel checks. Andreessen Horowitz, Thrive Capital, Accel, and Coatue led successive mega-rounds. Google and Nvidia are strategic investors in the Series D. And then there are the model providers — OpenAI (whose startup fund backed Cursor's seed) and Anthropic — whose APIs power the product while their own coding tools compete with it.

Truell rarely courts the press — but when he does appear, the conversations are worth reading.