The company reached a 50% enterprise revenue mix by mid-2026, with high-value customers now utilizing 8 times more tokens per user weekly than the average customer. CFO Sarah Friar reported that run-rate revenue for enterprise clients climbed 32% month over month in July. OpenAI is expanding into vertical AI applications for chip design, financial services, and life sciences to undercut open-source alternatives, such as GLM 5.3, on cost.
The vertical strategy includes the life-sciences model GPT Rosalind and the in-house Jalapeño chip, which completed its design in nine months before going to TSMC. OpenAI's advertising business reached a $1 billion run rate seven months after launch, the fastest any of its products have achieved that mark. Despite the growth, the company maintains a compute shortage that forces weekly trade-offs between training frontier models and serving users.
Key sources
- SOURCE@reuters“cost advantage over open-source”x.com
- SUPPORT@mtslive“specialized sectors, including chip design, life sciences and financial services”x.com
- SUPPORT@tbpn“the world's best healthcare platform”x.com
- SOURCE@tmtbreakout“Codex went from roughly 100,000 users at the start of the year to a last disclosed figure of about 25 million”x.com