OpenAI spent $34 billion last year, losses and revenue

OpenAI spent $34 billion in the past year. Independent journalist Ed Zitron compiled the figures, which were independently confirmed by the Financial Times.

The numbers

OpenAI's total outlay for the year came to $34 billion, far more than the year before. The largest single bucket was research and development, about $19 billion. Nearly $6 billion went to sales and marketing. Revenue for the period came in at roughly $13 billion.

By the end of the year monthly revenue hit $2 billion, up from $1 billion per quarter at the end of 2024. The company’s reported net loss jumped from $5 billion to around $39 billion.

How the loss breaks down

The Financial Times says most of the increase in the headline loss stems from a one-time, non-cash accounting charge of about $30 billion tied to OpenAI's earlier corporate structure. Strip that charge out and the loss sits at roughly $8 billion.

Those two figures create a split picture: on one hand there is continued heavy spending, especially on research and development; on the other hand there is accelerating revenue, with monthly receipts reaching $2 billion by year-end.

Why it matters

Heavy R&D spending shows OpenAI remains committed to development investment at scale, while revenue rising to $2 billion a month indicates growing commercial traction. The one-time non-cash $30 billion charge makes headline losses far larger than the company’s operating shortfall, which the data place at roughly $8 billion after adjustment.

That gap matters because investors and regulators looking at headline loss may draw a different conclusion than those who focus on recurring operating results. The company is preparing an IPO that could value it at more than $1 trillion, a context in which both the spending trajectory and the nature of accounting charges will be scrutinized.

What to watch

Watch whether monthly revenue keeps rising past $2 billion and how future filings and IPO documents disclose the $30 billion accounting charge and other non-recurring items. Those two signals will clarify whether the large headline loss reflects a temporary accounting anomaly or deeper structural spending above current revenue.