OpenAI has reportedly raised $11 billion at $1.25 trillion, a step up from March. But the round is smaller than planned, and Ramp data shows enterprise AI revenue growth stalling.
Photo by Tsuyoshi Kozu on Unsplash
OpenAI has reportedly raised $11 billion in a new round that values it at $1.25 trillion. Bloomberg broke the news, and it comes only months after the company raised $122 billion at $852 billion in March.
On the surface, that is investor appetite compounding. Look closer and the picture is more mixed.
The $11 billion is only part of a round that has not closed. Bloomberg's sources say it could reach about $20 billion in the coming weeks. In late September, the company was reportedly aiming for $30 billion at a $1.4 trillion valuation.
So OpenAI is landing below both its target size and its target price. The reported reasons are a drop in tech stocks and a sharp turn in public opinion against AI companies and data center projects, which the midterm results made visible. Investors are anticipating possible regulatory challenges ahead.
Anthropic's situation points the same way. It filed its S-1 for an IPO last week. Bloomberg reports that potential buyers are eyeing a valuation as low as $1.1 trillion, against the roughly $2 trillion investors were said to expect a month or so earlier. Other reporting puts the range at $1.2 trillion to $1.3 trillion. Either way, the number is far lower.
OpenAI, for its part, has pushed back its own IPO plans. Expectations now point to sometime next year.
The valuation is still going up. The cheque size and the revenue growth are not.
A new Ramp report helps explain the nerves. Companies are using more AI, but they are not paying more for it. Token spend on Ramp's platform grew just 1.5% across August and September, after growing 42% in July alone.
October brought a rebound of 8.9% between September 23 and October 23. Ramp's lead economist, Ara Kharazian, still says the headwinds are strong.
He does not blame weak demand. Volumes are rising. The pressure comes from price. OpenAI halved API prices for GPT-6 Sol and GPT-6 Luna in September. Anthropic then cut prices on Fable 5.1 and Opus 5.5, a move it said in a filing reduced its revenue growth by about $600 million.
Meta and Nvidia are pushing cheap or open models into the enterprise. Open-model makers in China and the US add to the pressure. So does Jev, a lightweight tool for picking between a few preset outputs, which Ramp says is increasingly among its top performers.
Customers are also getting smarter. Kharazian says mature users set default limits on costly models, cap employee budgets and route simple work to cheaper ones. His summary: "You can use more AI, but use it for less."
The spending is concentrated. Median firms spend $12 to $15 per employee per month, and the top 10% spend $600 to $1,000. The top 1% sit near $10,000 per employee. That group dipped slightly in late summer, and Kharazian says the dip alone is enough to stall overall revenue even as growth continues in the middle and upper tiers.
He doubts the rest of the economy will follow that cohort's lead. Coding agents have strong demand in software and finance, but a doctor's office or a car dealership may not need anything like that spend.
Our read: investors are not abandoning frontier AI, but they are repricing the growth story. A rising headline valuation sitting next to a shrunken raise suggests buyers still want exposure while questioning how fast revenue can climb.
The bull case rests on a new product category doing for spend what coding did. AI coding emerged about a year and a half ago and helped push both labs toward trillion-dollar valuations. Personal agents are the leading candidate, with OpenAI, SpaceXAI and Meta all shipping them. Kharazian is skeptical, since they are less economically productive than coding.
If that skepticism holds, the next test is not the size of OpenAI's round. It is whether usage growth can translate into revenue growth once prices stop falling. For now, the market is paying for the first and waiting to see the second.
Breakout
A year-old startup raised $45 million at a $300 million valuation to let lawyers, analysts and doctors retrain models themselves. Investors are backing a wider circle of AI builders.
Breakout
HackerTube turned years of Hacker News recommendations into a browsable video site. Within days, the same recipe was being applied to books and Wikipedia articles.
Breakout: China
Apple will bring a first set of AI features to Chinese iPhones with iOS 27.3. The slimmed-down launch, built around local partners, shows how AI products get adapted for a market.