OpenAI is looking like one of the most significant (and hyped) IPOs in years. But contrary to expectations, there are growing reasons to think it might not happen in 2026. The firm’s revenue miss will be forgotten in a quarter, but its cost architecture will not. The longer OpenAI waits to go public, the more likely it becomes that Anthropic and Databricks will set the valuation framework for the sector.

Whichever company defines the public market multiple for frontier AI gets to define what “good” looks like. OpenAI is at risk of walking into a comparison it did not set on terms it cannot control. That is the real cost of waiting.

Why OpenAI is likely not going public this year

  • A realistic IPO window has shifted from Q4 2026 to mid-to-late 2027. Public market investors will need to see how $1.15 trillion in infrastructure obligations convert into free cash flow. Such a large obligation requires several more quarters of clean execution before anyone can tell whether OpenAI is ready for an IPO.
  • The company’s infrastructure obligations are fixed, but its revenue is not. That mismatch is manageable when growth accelerates and existential when it does not. OpenAI does not manage this mismatch well.

  • CEO Sam Altman and CFO Sarah Friar’s joint denial of internal friction is more informative than the revenue miss. Strategic alignment on capital allocation is not something healthy pre-IPO companies need to publicly assert. When the CEO and CFO feel compelled to call a report “ridiculous,” the report is usually close to accurate.
  • Anthropic generates approximately $6.0 million in annualized revenue per employee, while OpenAI generates roughly $5.6 million. The gap is narrow today, but the trajectories diverge. Anthropic is scaling revenue faster than headcount. OpenAI is projecting significant headcount growth into a cost structure that is already under pressure.
  • Whichever company lists first defines the public market multiple for frontier AI. OpenAI is at risk of entering a valuation framework it did not set, on terms it cannot control, behind competitors that got there on a fraction of the capital. That is the highest-stakes consequence of waiting.

The IPO listing timeline is OpenAI’s first casualty

OpenAI missed multiple monthly revenue targets earlier this year after ceding share to Anthropic in coding and enterprise, and public market investors will want several more quarters of clean execution before anyone can credibly explain how $1.15 trillion in infrastructure obligations convert into positive free cash flow.

Mid-to-late 2027 is the more realistic window for OpenAI’s listing, which changes the AI IPO landscape. Anthropic and Databricks are further along in IPO readiness, according to our framework. If either company lists first, it will establish a public market comparable for frontier AI on materially cleaner unit economics. Being last to market after deploying the most capital is not a positioning any underwriter would choose. OpenAI’s fourth-quarter 2026 IPO target was already ambitious, given its corporate restructuring. In our view, it is now unrealistic.