How Cursor Accidentally Became a $60B Company

TThe Coding Koala
Computing/SoftwareSmall Business/StartupsBusiness NewsSpace/Astronomy

Transcript

00:00:00Okay, so this is the story of four guys who dropped out of MIT.
00:00:03They turned down offers from major big tech companies to build something of their own,
00:00:07and a year later, it completely flopped. And that failed project is the reason why
00:00:12Cursor exists today. This isn't your typical they struggled, then they became successful story.
00:00:17It's actually different than that, and you're going to love this one.
00:00:20Let's start with quick introductions. These are the four guys we will be talking about,
00:00:24and you don't really have to remember their names. You can just call the white guy number
00:00:27two in your head for the rest of this video. Just kidding. There's Michael Truel,
00:00:32who eventually becomes the CEO, and then Swala Asif, Arvid Lunmark, and Amon Sanger,
00:00:38all of whom met each other while studying computer science at MIT. And as I said earlier,
00:00:43they were dropouts. And even though they didn't graduate, they had received various job offers
00:00:47from big tech companies. So you can imagine how smart they were. Whereas people like us can't
00:00:53even find a job after graduating. Anyway, they rejected many job offers for the
00:00:57same reason they dropped out. To build something of their own. And as I said before,
00:01:02they didn't drop out to build Cursor first. They dropped out to build a CAD software. Which,
00:01:07in simple terms, is the software that lets engineers or architects build 2D or 3D models
00:01:13on their computer. They spent almost an entire year building tools for mechanical engineers,
00:01:18which again, none of them actually were. They were computer science guys trying to solve
00:01:22problems for an industry they'd basically only read about. And the thing is, it wasn't even a bad idea
00:01:28on paper. There genuinely was a gap in that market. But knowing a market has a gap and knowing how to
00:01:33actually fill a turnout to be two very different skills. And they didn't have a second one. They
00:01:37didn't understand the workflows. They didn't understand the pain points. They didn't understand
00:01:41any of the actual day-to-day misery of being a mechanical engineer. Truel himself later described this
00:01:47whole stretch as wandering in the desert. But if you've ever built something nobody wanted,
00:01:52you know exactly what he means. After wasting a year on that, they finally realized they had to do
00:01:58something in their own field and solve their own problems. And this is when the story gets
00:02:02interesting. And before I continue, I first want to introduce you to CodeCrafters. I assume many of
00:02:08us have gone from being software engineers to becoming prompt engineers, thanks to all these AI tools.
00:02:14CodeCrafters is a platform where you can get back those engineering skills again by building complex
00:02:19software from scratch yourself. It gives you guidance throughout the process, so you're not left
00:02:24staring at a blank editor wondering where to start. And once you complete these projects, you can even add
00:02:30them to your resume, which can help you stand out from the crowd. I've added a link in the description,
00:02:34so make sure to check it out. Now back to the video. They ditched the CAD thing, and now they were back to
00:02:40SquareOne, except this time they were actually thinking about what they're actually good at.
00:02:44More specifically, they were more drawn to the new trend of AI coding, things like GitHub Copilot,
00:02:50which most programmers were using, and which was fine, but these four looked at it and thought
00:02:54something everyone else didn't thought. Copilot wasn't actually part of your code editor. It was more
00:02:59like an add-on in VS Code for auto-completing your syntax. And that's when these four had the actual
00:03:04idea. What if instead of bolting AI onto an editor built for humans, you just built the editor itself
00:03:10around the AI from day one? And technically, doing that wasn't even some insane act of invention.
00:03:16They just took the open source code of VS Code, forked it, and started rebuilding it their own way.
00:03:21And they called their product Cursor, which if you think about it, isn't exactly the most creative
00:03:25name. Anyway, after the launch, their life was going to change in a very short amount of time.
00:03:30In 2023, Cursor was pulling in about $1 million in annual recurring revenue. And then in about 12
00:03:38months, that $1 million turned into $100 million, which is the kind of growth curve that doesn't
00:03:43really happen. And it happened so fast that people who study this stuff for a living started calling
00:03:48it one of the fastest revenue climbs any software company had ever pulled off. Faster than Slack,
00:03:53faster than Zoom, faster than basically anything you can name off the top of your head. And it didn't
00:03:58stop there either, because by the time we get to late 2025, Cursor had crossed $1 billion in annual
00:04:04revenue. And a big part of why this happened wasn't just that the product was good. It's that it landed
00:04:09at the exact moment the entire internet was falling in love with this idea of vibe coding. In simple
00:04:14words, they developed the right thing at the right time. So two months back in June of 2026, something
00:04:20happened that nobody saw coming. SpaceX. Yes, the Elon Musk company announced that it was acquiring
00:04:27AnySphere, the company behind Cursor, for $60 billion in an all-stock deal, which instantly made it the
00:04:34largest venture-backed startup acquisition that has ever happened. Not one of the largest, the largest.
00:04:39And if you're sitting there wondering what a rocket company wants with a coding tool, you're not alone,
00:04:43because SpaceX wanted to reduce how dependent it was on AI models built by other companies.
00:04:48And owning Cursor would give SpaceX direct access to a major AI developer tool platform
00:04:54and its engineering talent. The deal was closed recently in 14 August for $60 billion. It's great
00:05:00to see that, for dropouts, who once spent a year failing to sell software to mechanical engineers,
00:05:06just watched their second attempt get bought for $60 billion. And maybe that's really what this story
00:05:11is about. Not the money, not the acquisition, but the fact that failing at the first idea wasn't the end
00:05:17of anything. Cursor is a prime example of building the right product at the right time.
00:05:22Many people call it luck. Many people call it skill. Let me know what you think. Anyway,
00:05:27do let me know in the comments what you think about this whole story. And also make sure to check out
00:05:31CodeCrafters because it's more necessary than ever before. If you want to watch similar content,
00:05:37make sure to subscribe, and I'll see you guys in the next one.

Key Takeaway

AnySphere scaled Cursor from $1 million to $1 billion in annual recurring revenue in two years by building an AI-native code editor, resulting in a $60 billion acquisition by SpaceX.

Highlights

  • Four MIT computer science dropouts founded AnySphere after spending a year building unwanted mechanical engineering CAD software.

  • Cursor grew from $1 million in annual recurring revenue to $100 million in just 12 months.

  • Cursor crossed $1 billion in annual recurring revenue by late 2025.

  • SpaceX acquired AnySphere for $60 billion in an all-stock deal on August 14, 2026.

  • Cursor forks open-source VS Code code to build an editor designed natively around AI rather than adding AI as an add-on.

Timeline

Initial failures in computer-aided design

  • Four MIT computer science dropouts spent nearly a year building CAD software for mechanical engineers.
  • The founding team lacked understanding of mechanical engineering workflows and day-to-day user pain points.
  • Michael Truel described this initial stretch of building an unwanted product as wandering in the desert.

The founders turned down big tech offers to build their own software. They targeted a market they had only read about, resulting in a failed product that consumed an entire year before they pivoted to software engineering tools.

Pivoting to AI code editors

  • The team shifted focus to the rising trend of AI coding tools like GitHub Copilot.
  • Cursor forked open-source VS Code code to build an editor around AI from day one instead of treating AI as an add-on.
  • Annual recurring revenue reached $1 million in 2023.

Recognizing that existing tools bolted AI onto human-centric editors, the founders created Cursor by rebuilding the editor environment natively for AI-assisted workflows. This product launch quickly gained traction among programmers.

Rapid revenue growth and SpaceX acquisition

  • Annual recurring revenue climbed from $1 million to $100 million in 12 months.
  • Cursor surpassed $1 billion in annual revenue by late 2025.
  • SpaceX acquired AnySphere for $60 billion in an all-stock deal on August 14, 2026.

The rapid revenue climb outpaced major software benchmarks like Slack and Zoom, driven by the popularity of vibe coding. SpaceX purchased the parent company AnySphere to reduce reliance on external AI models and secure developer platform talent.

Community Posts

No posts yet. Be the first to write about this video!

Write about this video