In May 2026, just ahead of Y Combinator's Demo Day, roughly 180 startup founders received the same offer. It came directly from OpenAI CEO Sam Altman: instead of cash, OpenAI would provide API credits—tokens—and in return, each startup would hand over equity. The same terms, extended to every company in the batch.
TechCrunch called the scene a "mic drop moment." Like dropping the mic and walking off stage, the structure made acceptance feel more natural than pushback. Word is that YC founders spent several days debating the offer among themselves. For an early-stage startup, the chance to focus on building product without worrying about infrastructure costs is the kind of offer you have to prepare arguments to refuse.
If there is a reason to say no, what would it be?
The Moment Investment Became Compute
The structure of the deal is simple. OpenAI provides API credits instead of cash; the startup counts them as investment and gives up equity. The form of the investment shifted from money to computing resources.
The approach itself isn't new. AWS has been offering cloud credits to early-stage startups for years, and Google and Microsoft run similar programs. Treating services rather than cash as an investment vehicle is a strategy Silicon Valley has already validated.
But this case departs from precedent in a few ways.
What stands out first is scale: a single set of terms, offered to an entire YC batch at once. With one announcement, OpenAI built a structure that could place it in an equity relationship with 180 startups simultaneously. This isn't an investor assembling a portfolio one company at a time—it's binding an entire ecosystem in a single move.
The timing deserves a look, too. For startups building AI services, early infrastructure cost is one of the biggest barriers. The more a product depends on LLM APIs, the faster operating costs climb. Any team that has watched costs spike before traction arrives knows exactly how persuasive this offer's timing is. And what kind of negotiating position you hold when you show up with the solution at the precise moment the problem looks largest—that needs no explanation.
Above all, what makes this deal hard to read as a simple investment is who's making it. The CEO of the company that supplies AI infrastructure stepped forward himself, offering his own company's credits in exchange for startup equity. It's a scene where the language of an investor and the language of platform expansion merge into a single sentence. A structure like this only becomes possible when the company in question supplies essential infrastructure to the entire industry.
The Case for Skepticism
The view that this deal shouldn't be read in a purely positive light emerged right after the announcement.
The most direct criticism concerns dependency. A startup that architects its early service around the OpenAI API faces substantial costs to later switch to Anthropic, Google DeepMind, or open-source models. Codebase changes, prompt architecture redesign, response-handling rewrites, team retraining—all of these costs grow the more tightly the product is fitted to a single provider. The industry calls this switching cost. You start with free infrastructure, but you build—by your own hand, at the very beginning—a structure you can't exit when a better alternative comes along.
Then there's the asymmetry of valuation. In this deal, OpenAI sets the unit price of the credits, while the market sets the future value of the startup's equity. One side's value is fixed in the present; the other's is undetermined and lies in the future. Anyone who reads deal structures will quickly spot which side of that asymmetry holds the advantage.
There's another angle, too. API credits are a consumable resource, and they can only be spent on that one platform. Cash can go toward hiring, marketing, office space—whatever the moment requires. Credits can't. That's why some read this as trading equity, an asset with no restrictions on its use, for a resource whose use is restricted.
Of course, not every founder who took the offer signed without doing the analysis. From a startup's perspective, easing the burden of early infrastructure costs in order to focus on product can be a rational choice. What matters, I'd argue, is whether founders set their own criteria for calculating that rationality—or were simply persuaded by the appeal of the offer. The two processes can look identical in outcome, yet they produce entirely different positions at the negotiating table later on.
The Same Structure Has Already Reached Korea's Founders
This is a story from the American YC ecosystem. But for Korea's solo founders, one-person product managers, and practitioners adopting AI tools—and their counterparts everywhere—the same structure has already arrived in different forms.
AI SaaS products that pull you in with a free plan. Cloud partnerships that hand out free credits up front. Contracts that start at pilot pricing and adjust the rate later. Building your service with one specific AI provider's API as the default layer. Each of these is an inducement to form a dependency on a particular platform early. No one is asking for equity—but the logic of the trade isn't all that different.
Have you ever calculated the switching cost of the AI services you use today? Prompt redesign, workflow reconstruction, team retraining, data format conversion—add those costs up, and you'll find that no small number of organizations have already entered a serious state of dependency. The higher the switching cost climbs, the less room you have to respond when the subscription price goes up, the terms change, or the conditions of service shift.
It's also worth asking at what point a tool you use stopped being "a tool you could choose" and became "infrastructure you can't easily replace." Tools can be compared and selected; infrastructure is what you have once the cost of replacing it exceeds the cost of keeping it. Recognizing that boundary is itself the first step toward preserving future leverage.
A principle that recurs throughout the practice of sales and negotiation: when evaluating a deal, define the terms you can live with before you study the other side's offer. Move aggressively, but read the deal structure first—that habit is what keeps your judgment intact even in front of the most attractive proposal.
No one knows what kind of negotiating table the YC startups that accepted Altman's offer will be sitting at years from now. But the leverage they bring to that table depends, in large part, on how concretely they ran the numbers before saying yes today.
Receiving AI without cash changing hands is establishing itself as a new language of investment. Without the eyes to read that language, the most attractive offer can end up being the most expensive deal.
