OpenAI Enterprise Revenue Surpasses Consumer Business as Annualized Run Rate Hits $40 Billion

OpenAI's enterprise business has crossed a milestone, generating more revenue than its consumer-facing ChatGPT products, Chief Financial Officer Sarah Friar told investors on Friday — a shift that arrived ahead of the company's own internal projections.

"We entered the year at 60-40, but enterprise has accelerated much faster than expected and those lines have now crossed," Friar said, according to a person who attended the meeting but asked not to be named due to confidentiality. "The majority of our revenue is now enterprise."

Earlier this year, Friar had told investors that the company expected the consumer and enterprise businesses to reach parity by the end of 2026. The earlier-than-anticipated crossover signals an acceleration in adoption among business customers.

OpenAI's annualized revenue run rate has reached $40 billion, a figure that Bloomberg first reported and that was subsequently confirmed. The company's run rate grew 20% month over month in July, according to slides reviewed by attendees. Business customers expanded even faster, up 32% over the same period.

Friday's investor meeting was scheduled before a turbulent week in OpenAI's executive ranks. Revenue chief Denise Dresser stepped down Thursday, eight months after joining the company. Dresser, who spent more than a decade at Salesforce and served most recently as CEO of Slack, said in a LinkedIn post she was departing to pursue other opportunities.

Two days before Dresser's exit, longtime executive Brad Lightcap announced he was ending an eight-year tenure at the company to "start something new."

Greg Brockman, OpenAI's president and a co-founder, joined Friar for Friday's meeting. He thanked Dresser for her contributions and for building the enterprise foundation, the person said, and expressed enthusiasm about her replacement, Dali Rajic. Rajic previously served as operating chief at cybersecurity firm Wiz, now owned by Alphabet's Google. According to a separate person with knowledge of the matter, Rajic was introduced to OpenAI through Thrive Capital founder Josh Kushner.

Brockman also addressed questions about the competitive pressure from open-source Chinese AI models. He downplayed the threat and said there is a "misunderstanding" around open source being cheaper, the person in attendance said.

Friar described a behavioral shift among enterprise clients, framing it as a maturation of how companies deploy AI. The era of so-called tokenmaxxing — in which businesses allowed employees to accumulate large AI usage bills without demonstrating commensurate output — has ended, she said.

"Enterprise customers have moved from tokenmaxxing to focusing on cost per unit of intelligence," Friar said, according to the attendee.

She highlighted that OpenAI's newest model is "54% more efficient" on agentic coding tasks, and pointed to recent price reductions across the company's model suite as factors reinforcing that shift.

Friar also noted progress in advertising, telling attendees that OpenAI "made great strides" in that segment. The company's advertising business is approaching a $1 billion annualized run rate, according to the person present. OpenAI began testing ads within ChatGPT in February.

When executives were asked about the timing of a potential initial public offering, they declined to comment, citing a confidential SEC filing, the person said.

The session underscores the pressure on OpenAI's leadership to project stability even as key executives depart — and to demonstrate that the company's revenue trajectory remains intact ahead of what many analysts expect will eventually be a public market debut.

Disclaimer