OpenAI tops $10 billion in a single quarter for the first time, Bloomberg reports

OpenAI brought in more than $10 billion in revenue in the third quarter, the first time the ChatGPT maker has crossed that line in a single quarter, according to a Bloomberg report that cited people it did not name. The figures come from anonymous sources rather than a company disclosure, so they carry that caveat. If they hold up, they mark a sharp turn from earlier this year, when OpenAI's growth looked sluggish next to its main rival.

The jump is large. OpenAI reportedly recorded $6.7 billion in sales in the second quarter, so topping $10 billion three months later would mean an increase of almost 50%. Between the first and second quarters, revenue grew about 18%, from roughly $5.7 billion. Bloomberg attributed the acceleration to strong demand for the company's recent AI models.

An 18% quarterly gain on a base of more than $5 billion would be enviable for nearly any software company. Nearly 50% at this scale is unusual even by AI standards. It suggests OpenAI's newer models are giving paying customers something they were not getting before, rather than simply riding the popularity ChatGPT has enjoyed since its late-2022 debut.

The profitability picture improved too, though from a deep hole. OpenAI's adjusted operating margin narrowed to -109% in the third quarter from -150% in the second, Bloomberg reported. A margin below -100% means the company's adjusted operating loss was still larger than its revenue. With revenue above $10 billion, that implies a quarterly adjusted operating loss of more than $10 billion.

That is the number to watch. Revenue is rising faster than costs, which is the right direction. But TechCrunch's analysis credits part of the rebound to tighter pricing combined with discounts, alongside better models and a push into products for businesses and AI agents that carry out tasks for users. Discounts can win customers quickly, but they also make it harder to keep cutting a margin that is still deeply negative. A company that needs both rapid growth and a path to break even has less room to buy market share.

Bloomberg also reported that OpenAI's annual revenue run rate has reached $50 billion. Run rates project recent sales across a full year, so they flatter a company whose sales are rising fast. They do, however, show how much larger OpenAI's business is now than it was at the start of the year.

For all the growth, OpenAI is still chasing Anthropic. The Claude maker reported third-quarter revenue of $17 billion, about two-thirds more than OpenAI's reported figure. This week Anthropic said its annual revenue run rate had climbed by $10 billion in a single month, reaching $81 billion as of the end of October.

Those run-rate figures are not directly comparable. Anthropic counts sales made through its cloud partners, while OpenAI reports only its share of some sales that run through partners. The method gap almost certainly makes Anthropic's lead look larger than an apples-to-apples count would. Even so, the quarterly revenue gap of several billion dollars is hard to explain away with accounting choices alone.

The split comes down to customers. Anthropic has pulled ahead in revenue growth from enterprises, a market where OpenAI was slower to build products. OpenAI took a broader route and says it has more than a billion users. TechCrunch notes it is reasonable to assume most of them use the free tier or the cheapest paid plans. A huge consumer audience is valuable for brand and data, but business contracts are where the bigger checks are written. OpenAI's renewed focus on enterprise tools and agents looks like a direct response to where Anthropic has been winning.

The comparison carries extra weight now. Anthropic is seeking a $1.4 trillion valuation in an initial public offering that TechCrunch says is likely to happen soon. A listing would put its revenue and margins under quarterly public scrutiny and give investors a hard benchmark to hold OpenAI's privately reported numbers against.

Neither company is competing only with the other. Meta, Amazon, Microsoft and Google all offer their own models, and open-source labs, particularly in China, have released cheaper models that are still highly capable. That pressure helps explain why OpenAI has leaned on discounts. When a capable alternative costs much less, sticker prices are hard to defend.

The third-quarter figures, if accurate, show OpenAI can still grow quickly at a size few software companies ever reach. What they do not yet show is a business that makes money. Losing more than a dollar for every dollar of revenue, even on an adjusted basis, leaves OpenAI dependent on outside capital while a rival heads toward the public markets with sales roughly two-thirds larger. Closing the gap with Anthropic will take more than one strong quarter. The fourth quarter will show whether this one was a turn or a spike.