Sam Altman stunned the startup world this week with an offer that Y Combinator partner Tyler Bosmeny described as a “mic drop moment.” At a Y Combinator event on Tuesday night, Altman announced that OpenAI would invest $2 million worth of AI tokens into every startup in the current YC batch — not in cash, but in compute credits that startups can use to build their products. In exchange, each startup would give OpenAI equity through an uncapped SAFE note.
Y Combinator’s current cohort includes roughly 169 startups, according to its directory. For those companies, the offer represents a significant infusion of a critical resource: AI compute. Early-stage startups often burn through cash on infrastructure, and tokens—credits for using OpenAI’s models—can help them keep costs down while they iterate on product development.
The mechanics of the deal
The terms are structured as an uncapped SAFE (Simple Agreement for Future Equity), a standard instrument for YC’s early-stage deals. Unlike a traditional SAFE, an uncapped SAFE does not set a maximum valuation at conversion. That means the equity OpenAI receives will depend entirely on the startup’s valuation at its next priced round, typically a Series A. The higher the valuation, the smaller the slice of the company OpenAI gets.
Y Combinator managing director Jared Friedman confirmed the structure to TechCrunch. “It will convert in the next priced round, which is typically the Series A,” he said. This arrangement is designed to be founder-friendly — if a startup’s valuation soars before the conversion, OpenAI’s stake shrinks proportionally. Early estimates on social media suggest that if a startup reaches a $100 million valuation, OpenAI might end up with roughly 2% equity. Without seeing the formal documents, however, those figures remain speculative.
The token allocation itself is worth $2 million at current pricing, but the cost to OpenAI is lower and falling. As inference costs continue to drop — a trend driven by improvements in GPU efficiency and model optimization — the tokens given away today may cost OpenAI very little to produce tomorrow. This makes the equity OpenAI receives increasingly attractive from a cost perspective.
Strategic motivations for OpenAI
The deal serves OpenAI on two levels. First, it gains equity in a broad portfolio of early-stage companies. If even a handful of these startups succeed, OpenAI stands to profit significantly. Second, and perhaps more importantly, it locks these companies into the OpenAI ecosystem. Startups armed with $2 million of OpenAI tokens are likely to build their products on top of OpenAI’s models, defaulting to GPT rather than exploring competitors like Anthropic’s Claude Code or Google’s Gemini.
This kind of platform lock-in is a classic playbook in tech. Once a startup invests time and resources integrating with a specific API, switching costs become high. The tokens give a financial incentive to stay within the OpenAI orbit, potentially for years. For a company like OpenAI, which competes fiercely for developer mindshare, this is a powerful strategic move.
Founder reactions: pros and cons
Unsurprisingly, the offer sparked intense debate on X (formerly Twitter). Proponents argue that the deal solves a real pain point for early-stage startups: the skyrocketing cost of AI infrastructure. Compute can consume a disproportionate share of a young company’s budget, diverting money away from hiring, sales, and product development. By providing $2 million in tokens upfront, OpenAI gives founders breathing room to focus on building.
Critics, however, see significant risks. Seed investor Jason Calacanis, who runs a competing accelerator and fund, posted a stark warning: “If you take these tokens, there’s a non-zero chance that OpenAI will study exactly what your startup is doing, copy your idea and put your app into their free offering. This is the classic platform playbook — be careful, founders!” The fear that big AI labs could internalize the best ideas from startups is widespread and not unfounded. Both OpenAI and Anthropic have been known to release features that overlap with third-party applications built on their platforms.
Yet, as some commenters pointed out, this risk exists regardless of whether a startup takes equity or simply pays for tokens. Altman, as former head of Y Combinator and a regular guest speaker, already has access to YC’s demo days and pitch decks. The deal might even align OpenAI’s incentives with the startup’s success: if OpenAI holds equity, they have a financial reason to help the company grow rather than steal its thunder.
Equity dilution concerns
The bigger worry for many founders is equity dilution. Y Combinator already takes a standard 7% stake in exchange for $500,000 in cash. Seed investors frequently demand another 20% or more. Early employees need equity as compensation. Adding a third major equity holder — OpenAI — could leave founders with a smaller piece of the pie.
The actual dilution depends on the valuation at conversion. If a startup raises a Series A at a $50 million valuation, 2% to OpenAI is equivalent to about $1 million of equity value, which might be a fair price for $2 million in tokens. But if the startup fails to raise a high-priced round, or if it burns through tokens without reaching a liquidity event, the equity given away could be a poor trade.
Another danger: a startup could blow through its $2 million in tokens quickly without having built enough momentum to justify a subsequent round. In that case, the equity surrendered would have been for a short-lived advantage. Still, many founders believe taking tokens is better than paying cash, which is even scarcer at that stage.
Historical context: Altman and Y Combinator
Sam Altman’s ties to Y Combinator run deep. He served as president of Y Combinator from 2014 to 2019, overseeing its expansion into a global accelerator powerhouse. Under his leadership, YC invested in thousands of startups, including Airbnb, Stripe, and Dropbox. Altman’s own startup, Loopt, was a YC alum. His return as a benefactor and occasional speaker at YC events underscores his ongoing commitment to the startup community — though his primary role as CEO of OpenAI creates potential conflicts of interest.
OpenAI, originally founded as a non-profit research lab in 2015, transitioned to a capped-profit structure in 2019 to attract investment from Microsoft. The company has since become the dominant player in generative AI, with products like ChatGPT, GPT-4, and DALL-E reshaping industries from coding to creative writing. Its valuation has soared past $300 billion, making Altman one of the most influential figures in technology.
The token offer is not the first time Altman has used YC as a distribution channel for OpenAI’s services. During his tenure at YC, he frequently encouraged startups to use AI tools. But the current deal marks a more formal partnership, blending investment with product integration.
Broader implications for the startup ecosystem
The offer raises questions about the growing concentration of power in a few AI companies. If the largest AI players can dangle compute credits to capture equity in early-stage startups, they may end up controlling the next generation of innovation without having to generate ideas themselves. Critics warn this could stifle competition and lead to a handful of mega-corporations owning the AI infrastructure layer and the application layer.
On the other hand, the deal could democratize access to cutting-edge AI. Startups that cannot afford $2 million in compute suddenly get it for free (in exchange for future equity). This might allow more diverse founders to participate in the AI boom, leveling the playing field slightly.
Regulators are increasingly scrutinizing Big Tech’s investments in startups. Deals like OpenAI’s token-for-equity swap could be seen as a way to bypass antitrust rules by avoiding cash transfers. Whether this structure draws regulatory attention remains to be seen.
For now, Y Combinator’s current batch must decide whether to accept Altman’s offer. The window is likely short, as startups race to secure resources before the competitive landscape shifts. Those who decline may worry about falling behind peers who use OpenAI tokens to move faster. Those who accept must weigh the risks of platform dependency and dilution against the immediate relief of free compute.
As one founder on X put it: “Having $2M in free AI compute is like having a superpower — but it’s a superpower that comes with a pager always on.” The next few months will reveal how many startups choose to carry that pager.
Source: TechCrunch News