1. OpenAI revenue reaches about $10 billion for the quarter
OpenAI revenue reached about $10 billion in the third quarter of 2026, a rise of roughly 50% from the previous three months, according to figures that became public on November 10, 2026. The ChatGPT maker's annualized run rate stood near $50 billion at the end of September, and demand from business customers was a central driver of the growth.
The number places the company among a very small group of technology businesses that have reached an eleven-figure quarter while still privately held. It also shows how quickly the pace has changed within a single year. A 50% increase from one quarter to the next is far above what large software companies normally report, and it arrived on a base that was already measured in billions of dollars.
The timing adds to the significance. The quarter covers the three months from July through September, and the growth reported for it has been linked to demand from business customers rather than to any single event. For a company that sells access to AI models, rising revenue broadly tracks rising usage, so the figure is also a rough indicator of how much more work customers were sending to OpenAI's systems by the end of the quarter than at the beginning.
OpenAI is not a listed company and does not file financial statements with securities regulators, so its results do not come with the detail that public companies provide. The quarterly figure describes sales only. It says nothing on its own about costs, which for an AI developer include very large bills for computing capacity.
2. What is OpenAI's revenue run rate?
OpenAI's revenue run rate was about $50 billion at the end of September 2026. A run rate is an annualized figure: it takes the most recent period's sales and projects them across twelve months, on the assumption that the current pace simply continues.
The measure is popular with fast-growing companies because it captures momentum that a backward-looking annual total would miss. A business whose sales are climbing every month will always show a run rate higher than the revenue it actually recorded over the past year. That is useful for judging direction, but it is not the same as money already earned, and it can overstate results if growth slows or if a strong month proves to be an exception.
The run rate also helps explain how the quarterly and annual numbers fit together. Revenue of about $10 billion over three months averages a little more than $3.3 billion a month. A run rate near $50 billion implies that the pace at the very end of the quarter was higher than the average across it, which is what would be expected of a company whose sales were rising through the period.
3. Enterprise demand and the 70% figure
Sales to businesses sit at the center of the growth story. OpenAI sells to companies in two main ways: through subscriptions to its own products for employees, and through application programming interfaces, or APIs, which let other firms build its models into their own software and pay according to how much they use. Revenue of the second kind tends to rise with every additional product that a customer builds on the models.
Greg Brockman, an OpenAI co-founder, has said that the company's annualized revenue grew 70% from one quarter to the next. That figure is not the same measurement as the roughly 50% increase in quarterly revenue, and the two should not be read as conflicting. One compares total sales booked in the third quarter with total sales booked in the second. The other compares the annualized pace at the end of one quarter with the annualized pace at the end of the one before.
When growth accelerates late in a period, the second measure will come out higher than the first, because the end-of-quarter pace has pulled ahead of the quarter's average. The gap between 50% and 70% therefore points in the same direction as the run-rate arithmetic: the company was selling noticeably more at the end of September than it was at the start of July.
For a business whose customers are increasingly companies rather than individual subscribers, that pattern has consequences. Enterprise contracts are generally larger and last longer than consumer subscriptions, and usage-based API revenue can scale without a matching increase in the number of customers. Both features make revenue more predictable, which matters to the investors and partners financing the company's expansion.
4. How much Microsoft collects from OpenAI
Microsoft offers the clearest public window into OpenAI's finances, because it is a listed company with disclosure obligations and one of OpenAI's closest commercial partners. Microsoft said it recorded $10 billion from commercial engagements tied to OpenAI in its fiscal quarter that ended on September 30, 2026.
That amount is large even for a company of Microsoft's size. Total revenue for the quarter was $92.3 billion, which makes the OpenAI-linked portion almost 11% of everything Microsoft took in over those three months. Few single relationships account for such a share of a business that large.
The quarterly figure also represents a sharp step up from the recent past. Microsoft's revenue connected to OpenAI across the whole of its 2026 fiscal year, which ended on June 30, came to $24.1 billion. A single quarter at $10 billion is well above the average quarterly level implied by that annual total.
The two companies are bound together in several ways. OpenAI relies on Microsoft's Azure cloud to train and operate its models, and the firms have commercial arrangements that send money in Microsoft's direction as OpenAI's sales expand. For shareholders, the result is that Microsoft's reported numbers move with OpenAI's fortunes. An investor who cannot buy OpenAI stock directly can still track, through Microsoft's filings, how much business the AI developer is generating for its partner.
5. OpenAI revenue versus Anthropic
A record quarter did not put OpenAI at the top of its field. Anthropic, the developer of the Claude models and OpenAI's most direct rival, reported nearly $17 billion in revenue for the same three-month period. On third-quarter sales, that leaves Anthropic well ahead.
The comparison marks a notable change in how the two companies are viewed. OpenAI brought generative AI to a mass audience with ChatGPT and was long treated as the commercial leader by a wide margin. A quarter in which a competitor records sales roughly 70% higher complicates that assumption, at least on this one measure.
Some care is needed in setting the figures side by side. Revenue is only one indicator of a company's position, and the two businesses do not have identical customer bases or product lines. A quarterly total also does not capture the direction of travel in the way a run rate does. Even so, the size of the gap is too wide to be explained by definitions alone, and it has drawn attention away from what would otherwise have been a straightforward milestone for OpenAI.
6. Reading the numbers with care
Figures for private companies reach the public by a different route than those of listed firms, and that affects how much weight they can bear. A public company's revenue is audited, defined according to accounting standards and published in filings that anyone can read. A private company's numbers are shared with investors and partners and become known more widely only in part.
As a result, important context is often missing. The $10 billion total does not show how sales divide between consumer subscriptions, corporate plans and API usage, nor does it show what the company spent to earn them. Growth rates can be calculated in more than one way, as the 50% and 70% figures illustrate, and a reader who sees only one of them may draw a stronger or weaker conclusion than the full picture supports.
Microsoft's disclosures are valuable precisely because they do not share those limitations. They are prepared under public-company rules and cover a defined fiscal period. They describe only Microsoft's side of the relationship, however, so they confirm the scale of the commercial activity between the two firms without revealing OpenAI's own accounts.
7. What to watch next
The next reference point is the run rate at the end of the year. With the annualized pace at about $50 billion in September, the December figure will show whether the acceleration that Brockman described carried through the fourth quarter or eased.
Microsoft's next quarterly report will provide a second reading. If revenue tied to OpenAI holds at or above the $10 billion recorded in the September quarter, it would indicate that the step up from the fiscal 2026 average was not a one-time event. Because that revenue now approaches 11% of Microsoft's total, any change in it is material to the larger company's own results.
Readers following these updates should keep the two growth measures separate when new numbers appear. A fourth-quarter revenue total will be comparable with the roughly $10 billion booked in the third quarter, while a year-end run rate will be comparable with the $50 billion annualized pace recorded in September. Mixing the two would make growth look either faster or slower than it was.
The contest with Anthropic is the third element. Whether OpenAI narrows the distance from Anthropic's nearly $17 billion quarter, or the gap persists, will shape how customers, partners and investors rank the two companies heading into 2027.