OpenAI’s quarterly revenue reportedly crosses $10B for the first time

Bloomberg, citing anonymous sources, reports that OpenAI's quarterly revenue passed $10 billion for the first time. TechCrunch notes that would be an almost 50% jump from the roughly $6.7 billion reported for the second quarter, a sharp acceleration from the roughly 18% quarter-over-quarter growth reported earlier in the year.

The report also says OpenAI's adjusted operating margin improved from -150% to -109%, and that the company is at a $50 billion annual run rate. That lands a day after Anthropic said its own run rate reached $81 billion as it heads toward an IPO seeking a $1.4 trillion valuation. The race for who pays the bills in AI is now being scored in public.

We read this as a real turnaround, with an asterisk. A nearly 50% sequential jump at this scale doesn't happen by accident. TechCrunch points to tighter pricing, discounts, better models and a heavier tilt toward enterprise products and agents. That last part matters most to us. Customers paying for tools that finish actual work is the healthiest kind of AI revenue, and it's the clearest evidence yet that OpenAI is no longer leaning only on a huge consumer audience that mostly sits on free or cheap plans.

Now the asterisk. A -109% adjusted operating margin means the company is still spending far more than it brings in to deliver that revenue. Better than -150%, yes. But growth bought with discounts can look great for a quarter and then stall when the discounts end. I'm watching whether margin keeps narrowing as revenue grows, because that's what separates a business from a subsidy.

The Anthropic comparison also needs care. TechCrunch flags that the two companies count annualized revenue differently: Anthropic includes sales made through its cloud partners, while OpenAI reports only its share of some partner sales. So the headline gap between $50 billion and $81 billion is real in direction but probably overstated in size. Readers should resist treating these as like-for-like scoreboards, especially with an IPO prospectus about to shape the narrative.

Where does this lead if it scales? Two frontier labs pulling in this kind of money will set prices, terms and defaults for a large share of the software industry. That's useful when it funds better models. It's a concern when enterprise buyers get locked into one vendor's agents and pricing just as open-source and Chinese models are getting cheap and capable. Competition is the best protection buyers have, and it's still very much alive here.

One last caution: this is anonymously sourced and OpenAI hasn't confirmed it. We'd treat the direction as credible and the exact figures as provisional.

OpenAI commercialized large models early and says it has more than a billion users, but Anthropic pulled ahead in enterprise revenue growth this year and recently reported $17 billion in third-quarter revenue. Both also compete with Meta, Amazon, Microsoft, Google and a wave of cheaper open models.