TechCrunch reported that Sam Altman proposed at a Y Combinator event on Tuesday evening to offer each of this year’s YC startups a $2 million credit in OpenAI tokens in exchange for future equity. Based on YC’s roster, this year’s cohort includes approximately 169 companies.
This arrangement is not a cash investment, but rather provides startups with direct access to AI credits that can be used to develop their products. For OpenAI, this allows it to strategically position itself among early-stage companies while encouraging them to prioritize building on OpenAI’s models and tools.
Use an unlimited SAFE
YC Managing Director Jared Friedman told TechCrunch that the transaction will be structured as an "uncapped SAFE," meaning the agreement will convert into equity during the company’s next priced financing round, typically the Series A.
SAFE is a financing agreement commonly used by early-stage YC companies. The term "no valuation cap" means that no upper limit on the conversion valuation is set at the time of signing. The higher the startup’s future valuation, the smaller the equity percentage the investor typically receives upon conversion.
Therefore, startups currently cannot accurately determine how many shares to give up; the exact ratio will be determined only at the first priced financing round.
OpenAI simultaneously gains customers and equity holders.
The report highlights that this arrangement has two key implications for OpenAI. First, it allows OpenAI to acquire equity in a group of early-stage companies, enabling it to share in their future returns if they succeed. Second, the token allocation itself incentivizes these teams to prioritize OpenAI’s services over competing products like those from Anthropic.
The article also notes that as inference costs continue to decline, the credits OpenAI is giving away today may have an even lower actual cost in the future, making its transactions—exchanging equity in startups for these credits at a low marginal cost—appear even more favorable on paper.
Startups focus on cost and dilution
Supporters argue that this approach helps startup teams reduce one of the most easily inflated expenses: AI infrastructure bills. For early-stage companies with limited funding, these costs often significantly strain research and development and operational budgets.
However, opposition is also growing. Some investors worry that startups, after receiving allocations, may become even more dependent on a single AI platform; others caution that equity is equally scarce for early teams, as companies need to reserve space for future fundraising and employee incentives.
The report suggests that a more realistic risk is that startups may quickly exhaust their OpenAI credits without achieving sufficient business progress, while already facing equity dilution. However, from a cash flow perspective, for some early-stage teams, trading equity for compute credits may still be more palatable than paying cash upfront.