ANTHROPIC IPO COULD ECLIPSE EVERY SOFTWARE COMPANY IN HISTORY

Anthropic, the company behind the Claude family of AI models, has taken its first concrete step toward going public. The firm confidentially submitted a draft S-1 filing to the SEC on June 1, setting the stage for what market watchers expect to be the largest software IPO in history.

The timing is ambitious. Financial analysts expect the IPO to execute as early as fall 2026. By then, Anthropic's valuation will have nearly touched the trillion-dollar mark. Its most recent funding round, completed in May, valued the company at $965 billion.

you may like

IS AI THE NEW SPACE RACE? OPENAI IPO FILING KINDA CONFIRMS IT.

SPACEX BUYS CURSOR AFTER HISTORIC IPO

OPEN AI PREPARES PRICE CUT TO FIGHT ANTHROPIC

That figure puts Anthropic in territory only a handful of tech companies have ever reached while still private. It also means the early investors who wrote checks when the company was barely known are now sitting on stakes worth tens of billions. The question for public-market investors is whether the valuation can hold once the books are open.

THE $965 BILLION VALUATION IS NOT ASPIRATIONAL

It is the price at which investors bought in during the May raise, and it makes Anthropic roughly 18x the size of the largest software IPOs of the past decade. The last big one was Arm Holdings at $54 billion in 2023. Anthropic is seeking 18 times that.

The IPO itself will not sell at that full valuation. The S-1 process typically involves a price range set below the last private round to leave room for a pop. But the filing sets a ceiling, and the ceiling is higher than any software company has ever taken to market.

THE INVESTORS WHO ALREADY WON

The most direct beneficiaries of an Anthropic IPO are a handful of hyperscalers that started writing checks early. Amazon began investing in 2023 and has put in roughly $13 billion to date with plans for up to $20 billion more. By the time of the May raise, Amazon's stake was valued at roughly $74 billion according to an April regulatory filing. After the May round, that stake could exceed $100 billion.

Alphabet holds an estimated 14% stake. Based on the $965 billion valuation, that stake is worth roughly $135 billion. The relationship runs deeper than equity. Anthropic has committed to spending $200 billion on Google's cloud services and tensor processing unit (TPU) chips over the next five years. That is not a customer contract. That is an anchor lease on the entire Alphabet AI supply chain.

Nvidia invested up to $10 billion in a funding round late last year, a stake that is now worth significantly more after May. Microsoft also invested up to $5 billion in the same round, and has secured a $30 billion commitment from Anthropic for Azure cloud services.

Salesforce came in early with a $50 million investment in 2023, which has grown into a stake worth approximately $5 billion. That is a 100x return on a relatively small bet, the kind of gain that reshapes a corporate balance sheet.

THE CLOUD COMMITMENTS ARE THE REAL STORY

The equity stakes are impressive, but the cloud commitments are what make the valuation hard to argue with. Anthropic has basically pledged to run its entire compute load on the infrastructure of its largest investors. The $200 billion Google commitment plus the $30 billion Azure commitment equals $230 billion in guaranteed revenue for those cloud providers over the next five years.

That structure means the hyperscalers are not just passive shareholders. They are Anthropic's landlords. The equity returns are one payout stream. The cloud revenue is an entirely separate and recurring one. For Google and Microsoft, the IPO is almost irrelevant to the financial logic of the arrangement. The cloud contracts pay for themselves.

Anthropic also has a compute plan that matches the ambition of its valuation. It intends to deploy enough Nvidia Grace Blackwell and Vera Rubin chips to power an entire gigawatt of computing capacity. For context, that is roughly the same as a small nuclear reactor. No other AI lab has publicly committed to that scale of hardware procurement.

THE RISK FOR RETAIL INVESTORS

The valuation rests on a single assumption. That the demand for frontier AI models will continue to grow at a rate that justifies building a gigawatt-scale compute infrastructure. If that assumption holds, $965 billion might be cheap. If it does not, the cloud contracts become a fixed cost that no amount of equity dilution can escape.

There is also the question of concentration. The company's largest shareholders are also its largest customers and suppliers. That is not a problem for the IPO's success, but it is a structural risk for public investors who expect standard governance protections.

WHAT COMES NEXT!

The S-1 process typically takes 12 to 18 months. By the time Anthropic is ready to trade, the AI market will look different. There will be more competitors, more regulation, and potentially a macroeconomic environment that makes a $965 billion debut harder to sell.

But the filing is not a bluff. It is a signal that the company believes it can sustain the valuation through the public scrutiny that follows. The early investors have already won. The question is whether the rest of the market gets the same deal, or pays a premium for the privilege.

what to read next

MUSK IS A TRILLIONARE NOW, BUT HE STILL LIVES IN A 50K HOUSE

ANTHROPIC ACCUSES ALIBABA OF STEALING CLAUDE SECRETS

ANTHROPIC WANTS YOUR DRIVER'S LICENSE FOR SOME CLAUDE FEATURES

For now, the calendar says fall 2026. The paperwork says $965 billion. The rest is waiting.