The number circulating through the AI industry this week: OpenAI generated roughly $5.7 billion in revenue in the first quarter of 2026, according to The Information, citing two people with knowledge of the company’s financials. The figure tops rival Anthropic’s reported $4.8 billion for the same period by about $1 billion, and it suggests the company is on track toward the $30 billion annual target it has discussed with investors.
The drivers are the same ones OpenAI has been pushing for two years: Codex, the coding assistant that has become the company’s fastest-growing enterprise product; a broadening corporate sales effort; and early advertising tests inside ChatGPT. The ad experiments, in particular, are new money — OpenAI began offering cost-per-click campaigns in the consumer app, a revenue line that did not exist a year ago and one that scales with ChatGPT’s user base.
The cost of growth
The other number in the report is harder to celebrate. OpenAI’s adjusted operating margin was negative 122% in the first quarter, according to The Information — the company lost $1.22 for every dollar of revenue, even after stripping out stock-based compensation. The losses reflect the scale of its compute contracts and infrastructure spending, which have grown faster than revenue as the company races to secure capacity for its models.
That combination — fast-growing revenue and deeper losses — is the central tension of OpenAI’s path to public markets. The company is preparing an IPO that could value it near $1 trillion, according to people familiar with the matter, and investors will have to weigh whether the revenue trajectory justifies the burn rate. OpenAI’s position is that the losses are investment, not waste: the capacity being bought now is what makes the next generation of revenue possible.
The consumer engine
The consumer business remains the foundation. ChatGPT had about 55 million paying subscribers in the first quarter, up from roughly 47 million at the end of last year, according to people familiar with the figures. Weekly active users have climbed into the hundreds of millions, though the company missed its once-public target of crossing a billion weekly users by the end of 2025 — a miss that signaled the consumer market, while vast, is not infinite.
The composition of revenue is shifting, though. OpenAI’s chief revenue officer has said the commercial business is on track to account for half of total revenue by the end of the year, as enterprise deals for Codex and model access outpace consumer subscriptions. That mix change matters for the IPO story: enterprise revenue is stickier and higher-margin than consumer subscriptions, and it is the part of the business investors model most carefully.
The race underneath
The revenue disclosure lands in the middle of a two-front race. Against Anthropic, OpenAI is competing on absolute scale — more revenue, more users, more capacity — while Anthropic pushes toward profitability with a leaner cost structure and projects a jump to $10.9 billion in second-quarter revenue, according to investor materials reviewed by The Wall Street Journal. The comparison is stark: OpenAI is bigger today, but it is burning capital at a rate Anthropic has avoided.
Against the broader market, OpenAI is testing how much investors will pay for growth in a sector where the frontier keeps moving. The $5.7 billion quarter shows demand is real; the negative 122% margin shows the price of serving it. Between those two numbers sits the entire question of OpenAI’s valuation — and the answer will come when the company’s financials become public ahead of its listing.
The revenue mix is changing in ways that matter for the IPO. Advertising is the newest and fastest-growing piece — OpenAI began running cost-per-click campaigns inside ChatGPT, a market Google and Meta have dominated for two decades and one that monetizes attention rather than subscriptions. The company has also been integrating financial services into the app, connecting users’ accounts so the assistant can act on financial data, a step toward the platform ambitions that justify a trillion-dollar valuation. Each new revenue line adds surface area, and each one adds risk — advertising brings brand-safety questions, financial data brings regulatory scrutiny. The $5.7 billion quarter is the proof of concept; the S-1 will show whether the whole structure holds together.
For the public markets, the disclosure lands as a test of how AI financials are judged. OpenAI’s revenue is real and growing, but the negative margin invites the comparison investors have been making all year: revenue at hyperscale versus profitability at a premium. The company’s argument — that compute spending now is what makes future revenue possible — is the same one every cloud provider made in the last cycle, and it proved true for the biggest players and false for the rest. OpenAI’s S-1 will be the industry’s clearest case study yet of how that trade-off is priced.
OpenAI booked roughly $5.7 billion in first-quarter revenue, nearly $1 billion ahead of Anthropic, powered by Codex, enterprise sales and the first real advertising experiments in ChatGPT. The growth is genuine, and so is the cost: the company lost $1.22 for every dollar earned, a burn rate that defines the stakes of its coming IPO. Investors will be asked to price a company that is growing faster than almost anyone in the industry while spending faster still — and the market’s answer will set the template for every AI company that follows.
