US AI giant Anthropic said on 28 May (29 May AEST) it has raised US$65 billion at a post-money valuation of US$965, placing it ahead of rival OpenAI.
The new valuation follows the company’s Series H funding round, as it looks to expand computing capacity to meet growing demand for its Claude chatbot and scale its products.
By comparison, OpenAI was last valued at US$852 billion post-money in March.
It comes as both AI giants are expected to tap public markets, potentially within months, to secure the computing resources needed to power their services and train new models.
The so-called “AI IPO race” is shaping up as a reckoning for these very large companies that have remained private for years, and also comes as SpaceX moves closer to a public listing following its prospectus publication.
In recent months, the Claude chatbot creator has struggled to meet demand, forcing it to institute usage limits during peak hours and other measures.
This latest funding round was led by Altimeter Capital, Dragoneer, Greenoaks and Sequoia Capital, with Coatue and ICONIQ as co-leads, among others.
The round also drew participation from the firm’s strategic infrastructure partners, Micron, Samsung, and SK Hynix, and incorporates US$15 billion in previously pledged investments from hyperscalers, including US$5 billion from Amazon.
Speaking to Investor Daily, Global X senior investment strategist Billy Leung said Anthropic’s valuation surpassing OpenAI reflects less a verdict on who is leading the AI race, and more a broader shift in how markets are now pricing AI as a real revenue business.
“Anthropic’s run-rate revenue went from around US$1bn in early 2024 to over US$30bn by April this year, with enterprise adoption in code development carrying most of that growth,” Leung said.
“For both IPOs, the read-through is that public markets will be valuing these businesses on token economics and enterprise revenue rather than traditional software comparables.”
He contended that the broader AI boom has “clearly moved past” the question of who has the best model. Performance across the leading labs is now separated by weeks rather than quarters, and the cost of a given level of intelligence has fallen by orders of magnitude.
“The valuations are increasingly being justified by who can convert capability into recurring enterprise revenue, and that is also a signal for everything sitting underneath, since the next leg of the AI trade is increasingly about memory, advanced packaging and physical power rather than just the model layer.”
Earlier this week, Baillie Gifford investment analyst Kyle McEnery made a similar case to this publication, arguing that the “race” narrative obscures the bigger picture of a rapidly evolving AI ecosystem.
At the time, McEnery contended that both companies are well positioned to succeed, with their models already serving distinct purposes and operating in somewhat different domains.
“For OpenAI, they had the GPT moment, which has become one of the most memorable consumer applications, and as a result they ended up building out this consumer business.
“Anthropic initially started the other way around. They were very much focused on enterprise, and part of this was related to their founding cultural principles. They’ve always been very focused on AI safety.”
Anthropic’s new valuation also arrives alongside reports that SpaceX is targeting a valuation of at least US$1.8 trillion in its IPO, according to Bloomberg sources familiar with the matter.
Last month, the publication had reported that the rocket technology firm was aiming for a valuation above US$2 trillion, with the company reportedly settling on a lower figure following consultations with advisers and investors.
With CommSec recently naming itself the ‘lead Australian broker’ for the potential listing, local retail investor interest in the IPO is growing.
