OpenAI Tells Investors Revenue Is Closer to $50 Billion, Not $70 Billion

OpenAI has told investors its annualized revenue is nearing $50 billion, a figure that is roughly $20 billion below a recent estimate that had put the company closer to $70 billion. The revision matters because OpenAI is under intense pressure to show it can eventually justify the enormous sums being poured into its business.

The updated figure, reported by the Financial Times on Oct. 8, comes as OpenAI continues to burn through cash, delays a long-rumored public listing and faces growing scrutiny over how quickly the AI boom can translate into sustainable economics.

Only days earlier, reports circulating in the market suggested OpenAI’s annualized run rate had climbed near $70 billion. That number was widely discussed because it would have placed the company in a more direct revenue comparison with Anthropic, one of its biggest rivals. OpenAI’s latest disclosure to investors appears to pull that estimate back substantially.

The discrepancy highlights a larger problem in the fast-moving AI sector: there is no single standard for measuring revenue, and the headline numbers that investors and commentators quote are often based on different assumptions about what counts and when it should be counted.

What changed in OpenAI’s revenue picture?

OpenAI’s latest investor update indicates that its annualized revenue is closer to $50 billion, not $70 billion. In practical terms, that means the company’s business is still growing rapidly, but not quite as explosively as some earlier reporting suggested.

The difference is not just semantic. In a capital-intensive industry where companies are racing to build frontier models and scale infrastructure, a $20 billion gap can alter how outsiders view valuation, funding needs and the path to profitability.

The Financial Times reported that the earlier $70 billion figure had been derived from information previously shared by OpenAI with investors. According to the report, that estimate was also shaped by investor efforts to create a direct comparison with Anthropic’s annualized revenue.

Why does the $20 billion gap matter?

It matters because OpenAI is trying to persuade the market that its enormous spending is warranted by the size of the opportunity ahead. Revenue is the clearest evidence investors have that the company’s products can support the cost of training models, paying for compute and expanding globally.

Even with a $50 billion annualized run rate, OpenAI remains a business with extraordinary growth ambitions and equally extraordinary expenses. The lower estimate could sharpen questions about how quickly the company can narrow the gap between what it brings in and what it spends.

That issue has become more urgent since the company’s financials leaked earlier this year. Those documents reportedly showed OpenAI generated about $13 billion in revenue in 2025 while spending considerably more than it earned. For a company seeking to remain the defining leader in artificial intelligence, that imbalance is a central concern.

How OpenAI and Anthropic measure revenue differently

OpenAI and Anthropic do not appear to be using identical accounting conventions when they talk about annualized revenue. That difference makes direct comparisons risky, even when the headline figures seem similar.

According to the Financial Times, Anthropic includes sales made through cloud partners in its annualized figure, while OpenAI does not. That means Anthropic’s reported run rate may capture more of the commercial activity around its models than OpenAI’s does.

For investors, the distinction can be easy to miss but important to understand. Annualized revenue is often treated as a simple extrapolation from current monthly or quarterly performance. In AI, however, the growing use of channel partners, cloud platforms and enterprise resellers can make that extrapolation less straightforward.

OpenAI’s latest update appears to reflect a narrower way of counting revenue than some of the figures that have circulated publicly, while Anthropic’s own method reportedly includes cloud-partner sales.

Why do AI companies use different revenue methods?

They do so because the AI business is still evolving and because distribution increasingly runs through third parties. Some firms recognize revenue only from direct customer relationships, while others include partner-led transactions tied to their products.

That inconsistency can make one company look larger or smaller depending on the method used. It also complicates media coverage, since a “run rate” can sound precise even when the underlying definitions are not aligned.

What the revenue revision means for OpenAI’s funding story

OpenAI has raised staggering sums to fund model development, infrastructure buildout and product expansion. The company reportedly brought in $122 billion in a March financing round alone, underscoring how much capital investors are willing to commit to the platform.

Those investments are easier to justify if revenue is rising at breakneck speed. A number near $70 billion would have suggested an even faster path toward scale than a figure near $50 billion. While both are huge by startup standards, the difference changes the narrative around urgency and sustainability.

The company’s challenge is not simply to make money, but to do so while spending at a level that matches the frontier-model race. Training and serving large models is costly, and the infrastructure demands have only increased as users, developers and enterprise clients rely more heavily on AI systems.

OpenAI’s revenue numbers therefore sit at the center of a bigger question: can one of the most visible AI companies in the world eventually operate like a durable business, or will it remain dependent on continuous investment to support growth?

Why investors are paying such close attention

OpenAI has become more than a product company; it is a bellwether for the entire AI market. Its revenue growth, spending profile and valuation are watched as proxies for the health of the sector.

That makes every reported change in its financial trajectory consequential. A revision from $70 billion to $50 billion does not suggest a company in trouble, but it does temper expectations in an environment where hype often outruns disclosure.

It also raises practical questions about pricing, customer adoption and retention. To sustain a run rate of $50 billion, OpenAI must continue converting consumer enthusiasm and enterprise interest into recurring revenue at massive scale.

For a company with products used by millions, the challenge is not visibility. It is monetization. The market is now looking closely at whether subscriptions, API usage, enterprise contracts and emerging products can deliver the kind of cash flow needed to support its ambitions.

How does this affect OpenAI’s IPO timeline?

The revision does not change the fact that OpenAI’s IPO appears to be further away than many expected. Reports earlier this year suggested a public offering could materialize in 2026, but the company has since pushed that expectation back to early 2027.

That delay is consistent with the broader picture: OpenAI is still scaling, still absorbing immense capital and still working through the economics of an increasingly expensive AI race.

Going public too early could expose the company to heightened scrutiny before its financial model is fully mature. Waiting longer may give it more time to grow revenue, expand product lines and demonstrate that the business can sustain itself under public-market pressure.

What would a later IPO mean for the AI market?

It would likely reinforce the idea that the most valuable AI companies still need private capital to reach maturity. That, in turn, could encourage more large funding rounds, more reliance on strategic investors and more patience from the market about when profitability should arrive.

It could also suggest that OpenAI wants more time to resolve key questions around governance, spending and revenue recognition before being judged quarterly by public shareholders.

The broader context: a sector built on ambition and uncertainty

OpenAI’s latest revenue disclosure is a reminder that the generative AI market remains in a formative stage. Products have spread quickly, valuations have surged and investor enthusiasm has remained high. But the financial rules of the game are still being written.

Unlike mature software companies, frontier AI firms are balancing enormous infrastructure costs against business models that are still evolving. Revenue can rise rapidly, but so can spending on chips, cloud capacity and talent. That creates a moving target for profitability.

As a result, one of the most important stories in AI is not just which model performs best, but which company can turn technical leadership into a business that lasts. OpenAI’s revised revenue estimate is part of that story.

For now, the company remains one of the sector’s most powerful forces. But the latest numbers suggest investors should be careful about treating any single revenue figure as definitive in an industry where accounting methods, partnerships and growth assumptions can all shift the picture.

What happens next?

OpenAI has not publicly broken down the revised figure in detail, and TechCrunch said it reached out to the company for comment. Until the company provides more transparency, investors and observers are likely to continue debating which revenue number best reflects its real business momentum.

What is clear is that OpenAI is still growing fast, still spending heavily and still under pressure to prove that its massive market value is anchored in something more durable than hype. The revenue gap reported this week does not change the company’s central role in the AI boom, but it does make the financial picture a little less dazzling than some had assumed.

In a sector defined by dramatic claims and even more dramatic spending, that distinction matters.