OpenAI Revenue Crosses $10 Billion in a Single Quarter

Hammad Ali KhaksarHammad Ali Khaksar
November 12, 2026

“OpenAI's quarterly revenue has reportedly crossed $10 billion for the first time — a near 50% jump from Q2. Here is what drove it, what it means, and how it compares with Anthropic.”

OpenAI Revenue Crosses $10 Billion in a Single Quarter

OpenAI is losing money. A lot of money, by any normal measure. And it has just posted a quarter of revenue that most public companies would trade their entire year for.

That contradiction is the whole story of the AI business right now. According to Bloomberg, cited by TechCrunch on 11 November 2026, OpenAI's quarterly revenue crossed $10 billion for the first time in the third quarter of 2026. The figure is reported, not confirmed — Bloomberg's sources are anonymous and OpenAI has not published the number. But if it holds, it is a remarkable jump: $6.7 billion in the second quarter, $5.7 billion in the first. A near-50% rise in three months, after an 18% rise the quarter before.

The question is no longer whether people will pay for AI. It is whether anyone can make the maths work at this scale.

What the numbers actually say

The headline figure is $10 billion in a single quarter. Behind it sit three numbers that matter just as much.

First, the growth rate. Revenue rose from $5.7 billion in Q1 to $6.7 billion in Q2, then past $10 billion in Q3 — growth accelerating, not slowing.

Second, the annual run rate. Bloomberg puts OpenAI's annualised revenue at $50 billion. A run rate is not money already banked — it is what the current quarter's pace would produce over a full year — but it puts OpenAI among the largest software businesses on the planet.

Third, the losses. OpenAI's adjusted operating margin improved from minus 150% in Q2 to minus 109% in Q3. Read plainly: for every dollar of revenue, the company still spends more than two dollars running the business. The direction is better; the hole is still  enormous.

Why revenue jumped now

TechCrunch points to four drivers, and they fit together.

Better models brought users back and kept them paying. Tighter pricing and discounts did the quiet work — in a market this competitive, small pricing moves show up fast in the numbers. A sharper focus on enterprise products brought in customers that pay per seat, per month, and rarely leave. And AI agents — tools that complete tasks rather than just answer questions — are starting to turn ChatGPT from a product people use into one businesses depend on.

OpenAI spent the year rebuilding its commercial side after Anthropic started winning the enterprise argument. The $10 billion quarter is the first clear sign the rebuild is working.

OpenAI says it now has over a billion users, most of them free or on the cheapest plans. The challenge was never getting people through the door — it has been turning a billion curious users into enough paying ones to cover the cost of serving them.

OpenAI vs Anthropic: the revenue gap

The awkward part of the celebration is that OpenAI's biggest rival is ahead — and by a lot.

Anthropic reportedly recorded $17 billion in revenue in the same quarter, and said its annual revenue run rate rose by $10 billion within a single month, reaching $81 billion by the end of October. Anthropic is also heading for an IPO, reportedly seeking a valuation of $1.4 trillion. For context on that race, see our earlier report onOpenAI's $1.25 trillion valuation.

But the comparison comes with a warning label. The two companies count revenue differently: Anthropic includes sales made by its cloud partners, while OpenAI counts only its own share of some partner sales. So the $81 billion versus $50 billion gap is real, but not quite apples to apples.

Where Anthropic is clearly winning is the enterprise market, where companies buying AI for their whole workforce have been choosing Claude. OpenAI's response — enterprise products, agents, and brands operating inside ChatGPT — is aimed squarely at closing that gap.

Still losing money — and why nobody seems worried

A minus 109% operating margin would kill almost any other company. It has not killed OpenAI, for two reasons.

One is that the losses are, in a sense, the strategy. Training and running frontier models costs a fortune in computing power. Every dollar of compute buys capability, capability buys users, and users eventually buy subscriptions. Investors have funded the gap because the growth curve has never bent the wrong way for long.

The other is that the margin is moving in the right direction — from minus 150% to minus 109% in one quarter. If that trend holds for a few more quarters, the conversation changes completely: from "can this ever make money" to "when does it start making money."

But that "if" is doing heavy work: Meta, Amazon, Microsoft and Google are all pouring money into their own models, and cheaper open-source models — many from Chinese labs — keep pressure on prices from below. OpenAI's own cyber capabilities show how fast the frontier moves: its model recently found 948 Linux kernel exploits, as we reported in OpenAI's cyber model finds 948 Linux kernel exploits — the same capability race that drives revenue also drives costs.

What it means for everyday users

For the billion people using OpenAI's products, the $10 billion quarter changes little today and a lot over time.

In the short term, the free tier is safe — a company growing this fast does not need to squeeze free users tomorrow. In the longer term, it points to where OpenAI is heading: deeper into workplaces, shopping and services inside ChatGPT, and agents that do things rather than chat about them.

Worth watching next: whether Q4 keeps the growth rate up — one great quarter is a story, two is a trend. Second, what Anthropic's IPO does to the race, since a public Anthropic will have to show its numbers in daylight. And third, whether the margin keeps improving. Revenue is the headline. The margin is the business.

OpenAI crossed $10 billion in a quarter while still losing money on every dollar it makes. That sounds impossible. In the AI business in 2026, it is simply the price of staying in the race.

Enjoyed this guide?

Share it with fellow creators and developers.

Related Articles

View All Blogs →