Posts citing Financial Times reporting say OpenAI told investors its annualized revenue was approaching $50 billion at the end of September, about $20 billion below a figure that had circulated a month earlier. The summaries attribute the gap to different treatment of partner revenue.
A detailed account says investors adjusted OpenAI's July figure of about $30 billion upward to include estimated cloud-partner sales, making it easier to compare with Anthropic. They then applied OpenAI's reported growth rate to that adjusted base, arriving near $70 billion. The same post says the company's own September figure was just under $50 billion, still about $20 billion above July.
Two ways to count partner sales
The FT recap says Anthropic includes sales through AWS and Google Cloud while OpenAI does not. Another account of CNBC's reporting says the higher figure included revenue sharing from large partners, while the lower figure excluded those deals.
Those accounts describe different reporting scopes rather than two directly comparable measurements. OpenAI declined to comment to the Financial Times, according to the FT recap.
Why infrastructure investors care
The lower figure became part of a broader debate about AI spending. One market commentary argued that aggregate token growth matters more for data-center demand because chips, memory, networking and power can be used across several labs and enterprises. Other posts blamed the revenue headline for weakness in AI-linked stocks, but those are market interpretations rather than proof that one report caused the move.
The accounting dispute leaves OpenAI's exact comparable run rate contested in public discussion. It does not, on its own, settle how much demand the wider AI buildout will create.
