OpenAI’s revenue is reportedly $20 billion less than previously projected
Oct 8, 2026, 11:19 AM · TechCrunch

OpenAI's run rate is reportedly closer to $50 billion than $70 billion, and the bigger lesson is how loosely the AI industry's headline revenue numbers are being built and compared.
Why it matters
A little over a week ago, reports put OpenAI's annualized revenue near $70 billion, a figure that would have made it competitive with Anthropic's reported run rate. Now the Financial Times, as reported by TechCrunch, says OpenAI has told investors its annualized revenue is approaching $50 billion. That is roughly $20 billion lower.
According to the FT, the higher figure came from attempts by OpenAI's own investors to produce a direct comparison with Anthropic. That comparison is tricky because the two companies count differently: Anthropic includes sales made through its cloud partners, and OpenAI does not.
This lands while OpenAI is trying to justify enormous investment. TechCrunch notes the company raised $122 billion in a March round alone, that leaked 2025 financials showed about $13 billion in revenue against significantly higher spending, and that its IPO has been pushed to early 2027.
From the desk
Let's keep perspective first. A business approaching $50 billion in annualized revenue is enormous by any normal standard. Revenue at that scale means the demand for these tools is real. We do not read this as a collapse story.
What bothers us is how the bigger number came to exist. Per the FT, it was not OpenAI's own figure; it was investors trying to make OpenAI look comparable to a rival that counts revenue differently. That is a recipe for confusion. When the industry's two most-watched private companies use different definitions and the market fills the gap with adjusted estimates, the headline numbers stop meaning much at all.
And those numbers carry real weight. They anchor valuations, justify data center spending, and shape how lenders and suppliers size their own bets. A $20 billion swing in a reported run rate within about a week should make everyone downstream more careful, from chip suppliers to the cloud providers building for AI demand.
The IPO timing raises the stakes. Private investors can tolerate fuzzy, informally circulated figures. Public markets demand audited, consistent definitions. If OpenAI is heading to an early 2027 listing, the gap between the stories told in private and the numbers filed in public is exactly what gets scrutinized. Better that the correction happens now than in a prospectus.
Our read: OpenAI's business is large and growing, but the comparison game between labs has gotten ahead of the accounting. I'm watching for any standard way these companies start reporting revenue, and for whether other inflated comparisons quietly get walked back.
Context
Annualized revenue, or run rate, typically extrapolates a recent period's sales across a full year. It is a common metric for fast-growing companies but can be built in different ways, which is the heart of this discrepancy. TechCrunch said it had reached out to OpenAI for comment.
Who feels it
- Investors
- Run-rate comparisons between AI labs need a common definition before they can support valuation decisions.
- OpenAI
- The correction resets expectations ahead of a planned early 2027 IPO, where audited figures will replace informal ones.
- AI infrastructure suppliers
- Demand forecasts built on headline lab revenue deserve a second look when those figures can swing this much.
- Anthropic
- Its partner-inclusive counting method will draw more scrutiny as comparisons with OpenAI get sharper.
What to watch
- Any OpenAI comment confirming or clarifying the roughly $50 billion figure
- Whether OpenAI and Anthropic converge on comparable revenue definitions
- Updates on the timing and filing details of OpenAI's early 2027 IPO
- Investor reaction in OpenAI's next funding or secondary transactions
Companies: OpenAI