 |
| AI-generated illustration of a rising revenue chart with a dollar symbol, a laptop showing a chat AI interface, and a balance scale weighing gold coins against a stack of bills. |
On 10 November 2026, OpenAI told its investors something it had never been able to say before: it made $10 billion in a single quarter. Revenue for the third quarter of 2026 reached $10 billion, up nearly 50% from the previous three months, the first time the ChatGPT maker has crossed the ten-figure mark in one quarter. Chief financial officer Sarah Friar confirmed the figure in an interview on Tuesday, and the Wall Street Journal and The Information both reported the numbers from documents shared with investors. The growth is real, and it is fast: the second quarter grew only 18% over the first, and a year ago, in the third quarter of 2025, OpenAI reported sales of $3.9 billion. But the same documents carry the other half of the story, and it is just as striking. OpenAI's adjusted operating loss widened to $10.9 billion in the quarter, up from $10.1 billion in the second, which means the company spent more than a billion dollars more than it took in, even while growing at its fastest pace in a year. This article explains what the numbers actually say, where the growth came from, why the losses keep growing, and how the picture looks next to Anthropic, the rival that is about to go public.
The headline figures need a little unpacking, because OpenAI is a private company and these numbers come from investor documents rather than a public earnings release. Quarterly revenue of $10 billion takes OpenAI's annualised revenue run rate to roughly $50 billion at the end of September, according to Friar — the pace the business would reach if it kept earning at its current rate for a full year. The adjusted operating margin improved sharply on paper, from minus 150% in the second quarter to minus 109% in the third, which sounds like progress until you look at the loss itself: $10.9 billion, slightly higher than the $10.1 billion of the previous quarter. The reason the margin improved while the loss grew is that revenue grew faster than spending, so the loss shrank as a share of sales even as it grew in dollars. The word "adjusted" is doing real work here. These figures exclude stock-based compensation, which reached $2.6 billion in the third quarter alone, according to a person familiar with the matter cited by the Journal. Include that cost, and the true cash cost of running OpenAI is higher than the adjusted loss suggests. None of this is unusual for a company racing to build the infrastructure behind frontier AI, but it does mean the $10 billion headline and the $10.9 billion loss belong in the same sentence: OpenAI is now a very large business by revenue, and it is still a very expensive one to run.
As an Amazon Associate, we earn from qualifying purchases.
Prices and savings can change. Check the retailer for the current price.
Where the growth came from: enterprises doubled, consumers kept paying
The growth was not evenly spread, and the split between OpenAI's two businesses tells you where its strategy is working. Sales to business customers doubled from the previous quarter, Friar said, as the company's push into corporate accounts began to pay off. That enterprise surge is the main reason the quarter accelerated so sharply after the second quarter's comparatively modest 18% growth. New model releases helped, and so did price cuts: OpenAI has been cutting the prices of its latest models to catch up with Anthropic in the enterprise market, where its rival has been winning large corporate deals. The consumer side, which includes ChatGPT subscriptions, expanded about 20% from the previous quarter, a steadier but still strong pace. Roughly 100 million people now pay for ChatGPT, OpenAI told investors, equal to about 8.5% of its weekly active users — a slightly higher share than the previous quarter, suggesting the company is slowly getting better at converting free users into subscribers. ChatGPT now reaches 1.2 billion weekly active users, a number OpenAI has cited repeatedly this year, including at the launch of Businesses in ChatGPT on 10 November. Earlier this year OpenAI also rolled out advertising on ChatGPT, a move it expects to improve margins over time, though no figures for the advertising segment were available in these documents. Taken together, the picture is of a company whose consumer product is now a genuine mass-market subscription business — 100 million paying users is a scale few software companies ever reach — while its enterprise business is growing even faster from a smaller base.
The loss side of the ledger: why $10 billion in sales still loses money
The natural question after a $10 billion quarter is why the company still loses $10.9 billion in the same three months, and the answer is the cost of the AI race itself. Training and running frontier models requires enormous computing capacity, and OpenAI has committed to buying it at a scale that dwarfs its current revenue. The adjusted operating loss of $10.9 billion is the gap between what the company earns and what it spends on operations, before stock-based compensation — and that compensation is itself a major cost, at $2.6 billion in the quarter, paid to the researchers and engineers whose work the whole business depends on. The margin improvement, from minus 150% to minus 109%, shows the operating leverage working in OpenAI's favour: every extra dollar of revenue costs less than the last one to serve, because much of the infrastructure spending is fixed. But the absolute loss still rose, because the company is spending more in total, not less. This is the pattern of a company in a land grab: revenue growth of 50% in a quarter is the kind of number that justifies heavy spending to investors, as long as the growth continues. The risk is equally clear. If growth slows — if enterprise customers pause, if consumers stop converting to paid plans, or if a rival model wins the next cycle — the spending commitments do not shrink with it. OpenAI has not issued a comprehensive public quarterly financial statement of the kind a listed company must publish, so outsiders are reading these figures through investor documents and the reporting of the Journal, The Information and Bloomberg. What those documents show is a company that is now unambiguously big, and unambiguously unprofitable, at the same time.
The Anthropic comparison: $17 billion, and a much smaller loss
The most uncomfortable page in OpenAI's investor update is the one about its rival. Anthropic, the maker of Claude, reported third-quarter revenue of $17 billion for the quarter ended 30 September 2026, as disclosed in its securities filing ahead of its initial public offering — nearly three-quarters more than OpenAI's $10 billion. Its operating loss for the same period was $567 million, a fraction of OpenAI's $10.9 billion, and that figure includes stock-based compensation, whereas OpenAI's calculation leaves that cost out. The comparison is not perfectly clean: Anthropic's business is weighted far more heavily towards enterprise customers, while roughly half of OpenAI's revenue comes from consumers, and the two companies count some costs differently. But the direction is hard to argue with. Anthropic is earning substantially more and losing substantially less, and it is about to raise up to $64.4 billion in an IPO priced at $70 to $78 a share, with a separate $26 billion private placement, taking its total fundraising to as much as $90.4 billion. The timing sharpens the contrast: OpenAI shared its figures with investors on Tuesday, just ahead of Anthropic's long-awaited listing next week. The Journal's reporting notes that OpenAI has narrowed the gap with Anthropic in the third quarter, and the 50% growth rate is genuinely impressive, but narrowing a gap is not the same as closing it. For the first time, the AI race has a scoreboard that anyone can read, and on this quarter's numbers Anthropic is ahead on both revenue and efficiency.
What it means for the 1.2 billion people who use ChatGPT
For ordinary ChatGPT users, the financials matter less as numbers than as a guide to what the product will look like next. A company losing more than it earns has three ways to close the gap: grow revenue, cut costs, or find new revenue. OpenAI is visibly pursuing the first and the third. The growth is the enterprise push and the steady conversion of free users to the roughly 100 million paying subscribers. The new revenue is advertising, which arrived on ChatGPT earlier this year, and the commercial features launched this week: Businesses in ChatGPT, which lets companies claim verified profiles and run AI agents inside the chatbot, with sponsored placements as a destination for ChatGPT ads. The advertising business is the one to watch, because it is the lever that could change the product most. Ads inside a chatbot that 1.2 billion people use weekly are a large new revenue stream, and OpenAI has said it expects them to boost margins. The question for users is how visible that lever becomes: whether free users see more ads, whether paying subscribers stay ad-free, and whether business content starts to shape the answers themselves. OpenAI says claiming a business profile does not guarantee higher placement in ChatGPT's answers, and sponsored placements are labelled, but the commercial pressure is now explicit in the company's own numbers. The other change to watch is pricing. OpenAI has been cutting the prices of its latest models to compete with Anthropic for enterprise customers, which is good for developers in the short term, but a company that needs to fund $10.9 billion of quarterly losses cannot cut prices forever. None of this means ChatGPT is about to get worse or more expensive overnight. It does mean the product's economics are now visible, and users who understand them can read the next few product announcements — on ads, on pricing, on what stays free — with clearer eyes.
Checklist: what to watch after the $10 billion quarter
- Watch Anthropic's IPO next week: the listing will set a public market price on the rival's $17 billion quarter, and it will frame every comparison of the two companies from here on.
- Watch OpenAI's fourth quarter: the test is whether the 50% growth rate holds, or whether the third quarter was a one-off driven by enterprise deals and new model releases.
- Watch the advertising numbers: OpenAI has not disclosed what ads on ChatGPT earn, and that figure will show whether the free tier can pay for itself.
- Watch the paid subscriber count: 100 million paying users is a milestone, but the conversion rate is still only about 8.5% of weekly users, so there is a long way to go.
- Watch enterprise pricing: OpenAI cut prices to catch up with Anthropic, so the next question is whether it can hold those customers without cutting further.
- Watch the spending: the adjusted operating loss rose even as the margin improved, so the question is when total spending stops growing faster than revenue.
- Watch for a public filing: OpenAI is reportedly hoping to go public next year, and a listing would force the kind of detailed quarterly disclosure that private investor documents do not provide.
- If you pay for ChatGPT, check what your plan includes: the company's push into advertising and business features is aimed at the free tier first, but pricing and plan changes tend to follow the money.
Bottom line
OpenAI's third quarter is the best and the worst news the company has reported at the same time. Ten billion dollars of revenue in a quarter, up nearly 50%, with enterprise sales doubling and 100 million people now paying for ChatGPT, is the kind of growth that defines a generation of technology companies. An adjusted operating loss of $10.9 billion in the same quarter, with another $2.6 billion of stock-based compensation on top, is the kind of spending that defines a race nobody can afford to lose. The comparison with Anthropic — $17 billion in revenue, a $567 million loss, and an IPO next week — is the part of the story OpenAI would least like you to dwell on, and the part that matters most, because it shows the gap is still large and the rival is about to be priced by the public market. For users, the practical takeaway is simple: ChatGPT is now a mass-market product with a mass-market business behind it, and the company's next moves on advertising, pricing and paid plans will be shaped by the need to close a very large gap between what it earns and what it spends. The $10 billion quarter proves OpenAI can grow. The $10.9 billion loss proves it has not yet figured out how to grow profitably. The next two quarters will show which of those two facts matters more.
Reference sources
- The Wall Street Journal, via Israel Hayom, "OpenAI revenue jumps 50% to $10 billion as losses persist", 11 November 2026
- The Wall Street Journal, via Indian Television, "OpenAI revenue hits $10 billion in Q3 as operating loss widens to $10.9 billion", 11 November 2026
- The Wall Street Journal and Bloomberg, via Asiae, "OpenAI's Third-Quarter Revenue Surges 50%... Losses Widen, Too", 11 November 2026
- The Wall Street Journal and The Information, via AI Daily Post, "Anthropic Hits $17B as OpenAI Reaches $10B Revenue", 11 November 2026
- The Arabian Post, "OpenAI quarterly sales reach $10 billion as losses widen", 11 November 2026
- The Information, via Biggo Finance, "OpenAI's Quarterly Revenue Hits $10 Billion as Enterprise Sales Double", 11 November 2026
Leave a public comment