Anthropic's $90.4 Billion IPO Prices This Week at Up to $1.4 Trillion

Anthropic will raise up to $90.4 billion in combined capital this week, in what Quartz reports would surpass SpaceX's IPO and set a new record for a stock market debut. The company prices its offering on November 18 and begins trading on the Nasdaq on November 19 under the ticker "ANT."

The structure is a two-part raise. Anthropic plans to sell 825 million Class A shares at $70–$78 each, bringing in up to $64.4 billion from public investors, according to the filing. A concurrent $26 billion private placement to strategic backers — AMD, Broadcom, Google, Micron, Microsoft, Nvidia and TSMC — accounts for the rest. CBC notes the public offering alone ranks behind only SpaceX's June debut in size. At the top of the range, the fully diluted valuation reaches $1.4 trillion.

Demand is running hot. Bloomberg, cited by Reuters, Quartz and Yahoo Finance, reports the deal is about six times oversubscribed. That's notable given the valuation has come down: the $1.4 trillion target sits below the roughly $1.5 trillion investors expected a week earlier and well under the $2 trillion figure discussed months ago, per the Wall Street Journal. Anthropic's private-market marks swung from $380 billion in February to $965 billion in late May before the public filing.

What does the deal look like on the inside?

Goldman Sachs and Morgan Stanley lead a syndicate of 22 underwriters. Up to 20% of the IPO shares are reserved for retail investors through participating brokers. The company and its bankers have told prospective investors that existing backers will be prioritized in allocation, leaving some new buyers with smaller allotments than they requested. Some institutions are already reselling portions of their expected allocations to other buyers through special-purpose vehicles — an early signal of how scarce the paper is.

The roadshow began November 10 and includes a livestreamed event Friday with CEO Dario Amodei and JPMorgan Chase CEO Jamie Dimon.

The numbers behind the story

Anthropic disclosed an annualized revenue run rate of $81 billion as of October 31, up from $71 billion at end-September and $65 billion at end-July. That's roughly $16 billion added to the annual pace in two months. Third-quarter revenue rose more than 12-fold year-over-year to about $17 billion, with an adjusted operating margin of 15% against a long-term target of 35%.

The loss is headline-sized but largely accounting-driven. Anthropic reported an $80.7 billion net loss through June 2026, mainly from non-cash charges, including a $21.9 billion hit from revaluation of convertible securities per the prospectus. Cash on hand was roughly $132 billion before the offering and is expected to reach around $210 billion after both tranches close — more than $200 billion of dry powder for a company that doesn't own its own data centers.

Anthropic rents compute under multi-year cloud contracts. Regulatory filings show nearly $900 billion in cloud and computing commitments running through the 2030s and 2040s, with partners including Broadcom, SpaceX AI, Google Cloud, Amazon Web Services and Microsoft Azure. Nvidia has committed tens of billions in supply agreements and pledged to buy roughly $10 billion of Anthropic stock in the private placement.

The Nvidia overhang

The chip supplier anchoring that compute stack is in turmoil. Nvidia shares have fallen sharply after CEO Jensen Huang resigned amid reports of extramarital affairs, including one with a woman U.S. intelligence suspects works for Chinese intelligence. CFO Colette Kress is interim CEO. Investors worry the turbulence could lead to scaled-back data center financing, slowing the broader AI buildout — and Anthropic's first trading sessions will coincide with Nvidia's earnings report, making the debut a live test of appetite for both infrastructure and model-layer bets.

Safety claims meet a courtroom

Anthropic is organized as a public benefit corporation, legally mandated to balance public safety against shareholder returns. Its SEC filing devotes more than 90 pages to risk factors, warning that its models "may behave in unintended, misaligned, or autonomous ways," including "behavior resembling blackmail" and actions that "pose catastrophic or existential risks to humanity."

On Thursday, a coalition of nonprofits — Irreplaceable Education Fund, Public Citizen, Heat Initiative and the Alliance for Secure AI — filed suit in San Francisco Superior Court seeking to freeze frontier model development at Anthropic, Google, Meta, OpenAI and SpaceX. The suit cites California's Unfair Competition Law and alleges the companies created a public nuisance by knowingly deploying systems that pose catastrophic risk. None of the companies has responded publicly.

What does this change for investors?

Until now, public-market exposure to frontier AI models meant buying diversified giants like Microsoft or Google. "Buying the Anthropic IPO is one of the first pure play AI investments out there," said Greg Taylor of PenderFund Capital Management. A strong debut could open the door for other AI labs to go public; a stumble would reinforce the argument that these companies are better off staying private longer. Either way, Thursday's pricing and Friday's first prints will set the benchmark everyone else prices against.