OpenAI Q2 growth lags Anthropic as AI rivals near $100B run rate

OpenAI's second-quarter revenue grew more slowly than Anthropic's, disappointing some investors even as the ChatGPT maker said growth accelerated in the third quarter, widening a race toward $100 billion in combined annualized revenue.

"We entered the year at 60-40, but enterprise has accelerated much faster than expected and those lines have now crossed," Sarah Friar, OpenAI's chief financial officer, told investors at a closed-door meeting on Aug. 14, according to a person who attended. "The majority of our revenue is now enterprise."

Anthropic reported preliminary second-quarter revenue of more than $11.5 billion, up from $787 million a year earlier and $4.73 billion in the first quarter, according to documents seen by Bloomberg News. The Claude chatbot maker also posted its first quarter of positive adjusted operating income. OpenAI's annualized revenue run rate reached about $40 billion in July, doubling from roughly $20 billion at the end of 2025 in about eight months, with enterprise revenue growing 32 percent month over month.

Both companies have filed confidential IPO prospectuses with the Securities and Exchange Commission, and Anthropic could list as soon as this fall, before OpenAI. Together they are approaching $100 billion in combined annualized revenue, a figure that would have been implausible for the AI sector two years ago, and the divergence in growth rates is now a central question for investors weighing the two listings.

Enterprise crossover masks a slowing consumer base

OpenAI's structural milestone — enterprise revenue now exceeding consumer subscriptions for the first time — arrived two quarters ahead of its own forecast. As recently as October 2024, roughly 75 percent of OpenAI's revenue came from consumer subscriptions. The company passed 1 million paying business customers in November 2025 and now counts more than 2 million business customers and 1 billion weekly active users, roughly double a year earlier.

The enterprise engine runs on three products: the GPT-5.6 model series, the ChatGPT Work agent, and the Codex coding tool. Friar told the Aug. 14 meeting that the newest model is 54 percent more efficient on agentic coding tasks, and that recent price cuts — 20 percent on the mid-tier Terra model and 80 percent on the entry-level Luna — helped compete for cost-conscious developers. Advertising is emerging as a third pillar, approaching a $1 billion annualized run rate after testing began in February.

Anthropic's coding lead complicates the IPO math

Anthropic's growth is accelerating rather than plateauing: second-quarter revenue nearly doubled from the first quarter, and its annualized run rate crossed $47 billion in May, according to Bloomberg News, though the two companies calculate revenue differently. The surge is driven by Claude Code, which gained traction among developers before OpenAI's Codex response. OpenAI board chair Bret Taylor acknowledged the catch-up at a July all-hands, saying customers who "went deep on Claude Code" ended up with high bills and began looking for alternatives.

The competitive pressure lands as OpenAI manages an executive shake-up. Chief Revenue Officer Denise Dresser is leaving less than a year after joining, and Chief Operating Officer Brad Lightcap departed after eight years — both central to the enterprise business that now drives majority revenue. Dresser's replacement, Dali Rajic, previously served as president and chief operating officer of Wiz, the cybersecurity company Alphabet acquired for $32 billion.

The numbers carry a caveat: the $40 billion ARR is a run-rate projection, not audited revenue. OpenAI's most recent audited results, for fiscal 2025, showed a $20.92 billion operating loss against $13.07 billion in booked revenue, with $17.2 billion paid to Microsoft for Azure compute. Most analyst projections place breakeven no earlier than 2029. Anthropic, by contrast, reached positive adjusted operating income in the second quarter, though questions remain about whether the compute-cost timing that enabled the margin turn is sustainable. For investors, the two listings now offer a direct comparison: OpenAI with the larger absolute base but deeper losses, Anthropic with faster growth and a first profitable quarter — a divergence that will shape how the market prices both IPOs.

This article is for informational purposes only and does not constitute investment advice.