I totally get why OAI would want to have a presence at the application layer. Vertical integration and all that. But, like, they could so easily build it all in house. Take five engineers, give them each a million dollar salary, it's surely not going to take more than a year to build something comparable? There's no real tech moat there, which is why I originally found the sales price so insane.
From reporting it seemed like Windsurf had ~350 customers (enterprises). That's not nothing, but, come on. ~10M for CAQ? OpenAI has a pretty good brand! I would think they wouldn't find it all that hard to ink deals
I assumed they had millions of end-users. That makes an acquisition significantly less attractive. There's still the brand (various consumer good conglomerates acquire brands all the time) but without a strong user base the brand is worth significantly less. Confusing, then.
> Cognition did not announce the price it acquired Windsurf for; however, the company says Windsurf reached $82 million in ARR, with enterprise ARR doubling quarter-over-quarter. Cognition says Windsurf’s user base reached at least 350 enterprise customers and “hundreds of thousands” of daily active users.
I think you're allocating the $2.4B wrong. It went to Windsurf the company, not to the shareholders. If Windsurf paid out the employees, they had no obligation to pay out aquihires. They could have distributed severance to all the employees until $0 remained, leaving $0 to shareholders. So his better bet was to go with the rump Windsurf and the pile of cash.
I'm not sure I understand exactly what you're saying. I think you're saying, "Google just structured a bad deal for Prem and the other Windsurf employees, and Prem wanted money so he went back to Windsurf to get some of the additional cash reserves left there"
Just to note at the start, this whole thing is weird because technically, Google didn't even give the $2.4b to the company. Rather it gave the money to key employees and some investors. That is already weird, because this isn't actually an acquisition so much as a very aggressive hiring strategy.
I think your proposal (or at least what I understand of it) doesn't quite make sense. Windsurf wasn't liquidating the rest of its assets, it got acquired by Cognition. Prem (re)joined Windsurf after it had been acquired by Cognition. So if he was looking for liquidity or a cash payout, he wasn't going to get it at Cognition either.
And I also think this contradicts some of what we might expect of this deal. Yes, strictly speaking, it's NOT an acquisition, so Google could have given all $2.4b to the founders and told the rest of the key employees to kick rocks, join Google if they want, Google doesnt care. But that too seems weird. If the second employee wasn't worth actually trying to retain, what's even the point of this acquisition? What's Google getting out of this? Just the founders? No IP, no team, nothing else? It makes more sense that Google was trying to retain the key team, and the most reasonable way to pay them out is in proportion to the stock they already have.
I think what's weird about this whole saga is that the way folks are behaving doesn't line up with what I'd expect the incentives to be. Which means there's something in the way the deal is structured that doesn't make sense.
I think the licensing fee was paid to Windsurf, not employees or investors. The employees acquihired by Google got their own packages Maybe the total of all of the money going out from Google to everyone was $2.4 billion, not just the licensing fee.
And yes, they were acquired by Cognition, but when did the decision to go to Google explode?
Yup. It's confusing. I had to go back to the news stories to clarify myself too.
The thing is, Google can't just go ahead and give Windsurf's backers and non-aquihired employees money directly. What is it for? So they use a figleaf of a non-exclusive license to the technology. But it's a little bit too cute since according to Tan, it would have been enough to make everyone whole, including the investors and employees.
Although I wonder what happens to all that vested stock all the acquihired folks have.
As for joining Cognition, I guess it's worth it to Prem.
The incentive for OpenAI to purchase the application layer is it allows them to sell tokens at a 100x markup, as "credits" (Lovable).
I totally get why OAI would want to have a presence at the application layer. Vertical integration and all that. But, like, they could so easily build it all in house. Take five engineers, give them each a million dollar salary, it's surely not going to take more than a year to build something comparable? There's no real tech moat there, which is why I originally found the sales price so insane.
Acquire the userbase. TikTok would have gone nowhere if they hadn't been able to acquire musically and their 240M active users.
From reporting it seemed like Windsurf had ~350 customers (enterprises). That's not nothing, but, come on. ~10M for CAQ? OpenAI has a pretty good brand! I would think they wouldn't find it all that hard to ink deals
I assumed they had millions of end-users. That makes an acquisition significantly less attractive. There's still the brand (various consumer good conglomerates acquire brands all the time) but without a strong user base the brand is worth significantly less. Confusing, then.
Just checked and windsurf.com says they have over 1M users and 4k enterprise customers.
Hmmm someone isn't being truthful or is juicing numbers. This is from a few days ago from Cognition
https://techcrunch.com/2025/07/14/cognition-maker-of-the-ai-coding-agent-devin-acquires-windsurf
> Cognition did not announce the price it acquired Windsurf for; however, the company says Windsurf reached $82 million in ARR, with enterprise ARR doubling quarter-over-quarter. Cognition says Windsurf’s user base reached at least 350 enterprise customers and “hundreds of thousands” of daily active users.
I think you're allocating the $2.4B wrong. It went to Windsurf the company, not to the shareholders. If Windsurf paid out the employees, they had no obligation to pay out aquihires. They could have distributed severance to all the employees until $0 remained, leaving $0 to shareholders. So his better bet was to go with the rump Windsurf and the pile of cash.
I'm not sure I understand exactly what you're saying. I think you're saying, "Google just structured a bad deal for Prem and the other Windsurf employees, and Prem wanted money so he went back to Windsurf to get some of the additional cash reserves left there"
Just to note at the start, this whole thing is weird because technically, Google didn't even give the $2.4b to the company. Rather it gave the money to key employees and some investors. That is already weird, because this isn't actually an acquisition so much as a very aggressive hiring strategy.
I think your proposal (or at least what I understand of it) doesn't quite make sense. Windsurf wasn't liquidating the rest of its assets, it got acquired by Cognition. Prem (re)joined Windsurf after it had been acquired by Cognition. So if he was looking for liquidity or a cash payout, he wasn't going to get it at Cognition either.
And I also think this contradicts some of what we might expect of this deal. Yes, strictly speaking, it's NOT an acquisition, so Google could have given all $2.4b to the founders and told the rest of the key employees to kick rocks, join Google if they want, Google doesnt care. But that too seems weird. If the second employee wasn't worth actually trying to retain, what's even the point of this acquisition? What's Google getting out of this? Just the founders? No IP, no team, nothing else? It makes more sense that Google was trying to retain the key team, and the most reasonable way to pay them out is in proportion to the stock they already have.
I think what's weird about this whole saga is that the way folks are behaving doesn't line up with what I'd expect the incentives to be. Which means there's something in the way the deal is structured that doesn't make sense.
I think the licensing fee was paid to Windsurf, not employees or investors. The employees acquihired by Google got their own packages Maybe the total of all of the money going out from Google to everyone was $2.4 billion, not just the licensing fee.
And yes, they were acquired by Cognition, but when did the decision to go to Google explode?
O I see what you're saying. The wording is actually very unclear -- for example the WSJ says $2.4b to license AND hire the CEO (https://www.wsj.com/tech/ai/google-to-pay-2-4-billion-in-deal-to-license-tech-of-coding-startup-hire-ceo-b9b94bbc) while Reuters says licensing fee and investor liquidity (https://www.reuters.com/business/google-hires-windsurf-ceo-researchers-advance-ai-ambitions-2025-07-11/). It would be odd if Google paid $2.4b just to the company though -- why would Varun (the CEO) say yes to a deal where he gets nothing? Probably the $2.4b is a mix of buying out the founder and investors directly, plus some to license the tech.
IDK about the specific timing but Prem explicitly says he's joining Cognition in his tweet
Yup. It's confusing. I had to go back to the news stories to clarify myself too.
The thing is, Google can't just go ahead and give Windsurf's backers and non-aquihired employees money directly. What is it for? So they use a figleaf of a non-exclusive license to the technology. But it's a little bit too cute since according to Tan, it would have been enough to make everyone whole, including the investors and employees.
Although I wonder what happens to all that vested stock all the acquihired folks have.
As for joining Cognition, I guess it's worth it to Prem.