OpenAI annualized revenue reported at $50bn not $70bn

By Splitfeed EditorialLast updated Oct 8, 2026, 18:31 UTC

On October 8, 2026, the Financial Times reported that OpenAI's annualized revenue stands at approximately $50 billion, roughly $20 billion below figures that had previously been signaled to investors. Other outlets confirmed the gap, noting that the earlier figure widely circulated among investors and analysts had been around $70 billion. OpenAI did not publicly confirm or dispute the revised number. The disclosure moved markets: shares of Nvidia, Oracle, AMD, and CoreWeave fell following the report, and the US100 index dropped approximately 2 percent on the day.

The significance runs beyond a single data point. OpenAI is in the middle of an aggressive infrastructure and compute spending cycle, and the valuation attached to the company depends heavily on expectations about revenue trajectory. A $20 billion gap between what investors were told and what the Financial Times is now reporting raises questions about how AI lab revenues are measured, how reliably those figures are communicated, and whether the growth curve assumed by the market is materializing. The moment is particularly sensitive because OpenAI recently completed a large funding round at a valuation premised on continued steep revenue growth.

The anti-AI camp is treating the revision as a meaningful signal of trouble. Dare Obasanjo noted that he had long wondered when an AI lab would be hurt by treating usage-based revenues as subscription revenues, since the former can fall. An account called WildPartyInviteDecliner argued the revised figure implies OpenAI has barely grown revenues since summer. Don Pittis asked whether this is the turning point many have been worried about, and Sabstuck tied the figures to broader concerns, pointing to Anthropic's reported $8 billion in losses against $4.6 billion in revenue as additional context. The middle ground camp is more cautious. Cape Fear Advisors argued the gap is largely definitional, with Anthropic counting sales through cloud partners while OpenAI does not. Robert Lee called it an accounting issue, and Librarian Capital suggested outside analysts applied a correct growth percentage to the wrong base figure. David Crespo added that the differing reporting methodologies were already understood and that few were placing heavy weight on the specific $70 billion number. The pro-AI camp has not yet published reactions, though camps with that orientation would typically argue that current revenue figures understate the longer trajectory and that platform adoption compounds in ways early snapshots do not capture.

The clearest resolution would come from OpenAI directly confirming or clarifying its annualized revenue figure and explaining how the $70 billion number entered circulation. Any investor update, formal earnings disclosure, or regulatory filing tied to OpenAI's ongoing corporate restructuring would be the next concrete data point to watch.