OpenAI Revenue Restated: Twenty Billion Less Than Reported

OpenAI Revenue stood at about fifty billion dollars on an annualized basis at the end of September, not the nearly seventy billion dollars that circulated through financial media late last month. The corrected figure comes from documents the company shared with investors, as reported by the Financial Times on October 8. The gap rattled markets: the Nasdaq Composite fell more than a percent on Thursday, its worst session since mid-August, and chip and infrastructure stocks tied to the AI boom slid with it.

Here is the twist that makes this more than a story about one number moving. The missing twenty billion dollars of OpenAI Revenue was never lost, stolen, or even really there. It was a construction — an attempt by OpenAI's own backers to produce a tidy comparison with Anthropic, the rival whose annualized revenue had passed sixty-five billion dollars by July, according to Reuters. When OpenAI told investors the true figure, disclosed through the Financial Times, the machines that make the machines trembled.

How OpenAI Revenue gets counted

Annualized revenue run rate is the startup world's favorite magnifying glass. Take one month of sales, multiply by twelve, and project what the year would look like if the current pace held. It is a pace, not a pile of money actually earned — and as Mint's breakdown of the Financial Times reporting explains, that is how OpenAI Revenue can swing by tens of billions of dollars without a single customer leaving.

The swing here comes down to cloud partners. Anthropic includes revenue from sales made through cloud providers such as Amazon Web Services and Google Cloud when it reports its run rate, then records the provider's cut as an expense. OpenAI Revenue only counts the company's own share of those partner deals, as reported by the Financial Times and detailed further by The Decoder, citing Axios. Both methods comply with American accounting standards, according to Axios, and which one a company uses depends on who controls the customer relationship.

The higher figure — roughly sixty-eight billion dollars — was designed to make OpenAI Revenue directly comparable to Anthropic's number, Axios reports. Strip away that comparability exercise and the company is at about fifty billion, a number that tells a calmer story about very real growth: OpenAI started this year at roughly twenty billion in annualized revenue versus about six billion in twenty twenty-four.

The rivalry behind the accounting is getting louder. For the first time, Anthropic eclipsed OpenAI on quarterly revenue in the second quarter — eleven and a half billion dollars against six point seven billion, according to Mint. Anthropic's annualized run rate crossed sixty-five billion in July and sources told Reuters it could reach a hundred billion by year-end. OpenAI, meanwhile, says its overall annualized revenue grew seventy-seven percent during the third quarter, with enterprise sales jumping more than a hundred percent, CNBC reported.

Why the market flinched anyway

If this was just a measurement story, traders would have shrugged. They did the opposite. Shares in Nvidia fell around three percent, Oracle six percent, and CoreWeave eight percent in intraday trading, according to StartupFox. Advanced Micro Devices and Broadcom each dropped about five percent, with Intel and Super Micro Computer down roughly six percent. One revised cell in a spreadsheet, and the wider AI infrastructure complex shuddered.

The nervousness has a price tag attached. OpenAI is negotiating at least thirty billion dollars in new capital at a target valuation of one point four trillion dollars before the money goes in, Bloomberg reported, after raising up to a hundred and twenty-two billion dollars in March at a post-money valuation of eight hundred and fifty-two billion. The company recorded a net loss of thirty-eight point five billion dollars in twenty twenty-five on about thirteen billion in revenue, according to StartupFox. At DevDay, the company put its scale on display: more than one point two billion weekly ChatGPT users and two and a half million businesses on its products, The Decoder reported.

The company insists OpenAI Revenue will keep climbing, targeting at least a seventy-billion-dollar run rate by the end of twenty twenty-six, Bloomberg reported, driven by enterprise adoption. But the selloff shows how tightly markets track the two biggest American AI companies — and how jittery investors have become about the AI trade, with OpenAI's public listing now pushed to twenty twenty-seven and Anthropic's own IPO reportedly moving cautiously, The Decoder noted.

What OpenAI Revenue means for everyone watching AI

There is a lesson here for anyone who does not trade stocks for a living. When two companies chasing the same customers count revenue differently, the headline numbers stop being comparable — and both have now filed confidential paperwork to go public, according to Mint. Anthropic could list as early as this fall, while OpenAI says it no longer expects to go public this year. A listing would give ordinary investors their first real look under the hood of companies whose valuations currently run on trust.

The bigger question is the one feeding the AI bubble debate: can revenue growth at AI companies keep pace with their enormous spending on compute and data centers over the long run? The Decoder frames the answer as hinging on measurable productivity gains that businesses can actually point to — the difference between AI as a useful tool and AI as an expensive habit.

For a generation entering the workforce as AI reshapes hiring, wages, and entire job categories, that distinction matters more than any single quarter. A twenty-billion-dollar gap in OpenAI Revenue turned out to be an accounting choice, not a collapse in demand. The next time a giant number about AI lands in your feed, the right question is not whether it is big. It is how it was counted — and who benefits from the comparison.

More AI News | Related: Anthropic Says Stop Being Cruel to Claude

Sources: Mint, reporting on the Financial Times, The Decoder, citing Axios, Bloomberg, and CNBC, and StartupFox on the market reaction.