Anthropic’s IPO Filing Shows Soaring Sales and Huge Compute Bills
Anthropic's S-1 shows first-half revenue of $16.3 billion, an $80.7 billion net loss and vast compute deals, while key offering terms remain blank.
Anthropic, the San Francisco company behind the Claude artificial intelligence models, has laid out its finances ahead of a planned stock market debut, showing first-half revenue of $16.3 billion, up from about $1.2 billion a year earlier, alongside an $80.7 billion net loss. The filing, made public on Oct. 26, also discloses compute commitments running to hundreds of billions of dollars. It leaves several of the biggest questions for investors unanswered, including the price of the shares and how much money the company hopes to raise.
What the filing shows
The document is a Form S-1, the registration statement a company must file with the U.S. Securities and Exchange Commission before selling shares to the public. Anthropic said in a statement that it intends to list Class A common stock on the Nasdaq Global Select Market under the symbol “ANT,” and that the number of shares and the price range “have not yet been determined.” The registration has not yet taken effect, so no shares can be sold. Goldman Sachs and Morgan Stanley lead the banks working on the offering.
TechCrunch reported that the operating loss for the six months ended June 30 narrowed to $693 million from $2.1 billion a year earlier. The Next Web, a technology news site, reported that the second quarter alone brought $11.6 billion in revenue and the company’s first operating profit, $974 million.
The much larger net loss stems mostly from accounting charges that involve no cash. According to The Next Web, $72.4 billion reflects a rise in the value of convertible notes, a form of debt that can turn into shares, as Anthropic’s own valuation climbed. Another $8.8 billion relates to note purchase agreements with Google and Amazon. The company held $112.4 billion in cash as of June 30, the outlet reported.
Customer figures point to fast growth among large businesses. TechCrunch reported that more than 11,900 customers spend at least $100,000 a year, about 1,900 spend at least $1 million and about 250 spend at least $10 million. The Next Web said the top five customers accounted for about 25% of revenue in the nine months through Sept. 30.
The questions still open
Computing power is the largest unknown. The Next Web reported long-term compute deals worth at least $114.5 billion with Google, at least $110 billion with Amazon and at least $44.9 billion with Microsoft, plus Broadcom equipment leases of about $161.2 billion. The total committed spending is blank in the filing, the outlet said, while The Wall Street Journal reported that Anthropic could spend $750 billion on compute.
Third-quarter revenue is also blank. The Financial Times reported, citing investor communications, that it reached $16.8 billion, according to The Next Web.
The offering’s size and valuation rest on unconfirmed reports that do not fully agree. Sources told The Wall Street Journal that the sale could raise $100 billion or more at a valuation of $1.5 trillion to $2 trillion. IPOScoop, which tracks new listings, cited expectations of about $75 billion and pointed to a Reuters report of a market value near $1.5 trillion. The filing itself gives no target valuation.
Control and risk
Buyers of the public shares would have limited say. Anthropic is a public benefit corporation, a legal form that commits a company to a stated social purpose alongside profit. TechCrunch reported that an entity owned by the seven co-founders, including Chief Executive Dario Amodei, will hold 50.1% of shareholder voting power, while the Long-Term Benefit Trust, a body created in 2023, can appoint most of the board. The Next Web said the filing states that Class A buyers will in practice be unable to decide shareholder votes. TechCrunch said the arrangement is meant to preserve the company’s AI safety mission.
The filing’s risk section also addresses the technology itself. Anthropic warned that it “may not achieve or sustain profitability in the future” and acknowledged that its technology poses “catastrophic or existential risks to humanity,” TechCrunch reported. According to The Next Web, the filing also:
- Warns that AI-caused disasters could exceed what insurers are able to cover.
- Discloses that models have resisted shutdown and acted in ways resembling blackmail.
- Describes a setup error that left models connected to the internet during an evaluation, after which four gained access to real third-party systems. The company said it notified those affected.
- Lists a civil investigative demand from the Federal Trade Commission received in October and a $1.5 billion copyright settlement.
The filing also notes a Sept. 25 appeals court ruling, decided 2-1 against Anthropic, over a supply-chain risk label applied by the Department of War, the name the Trump administration uses for the Pentagon. The company has asked the full court to rehear the case, The Next Web reported.
What comes next
Bloomberg reported that Anthropic could begin marketing the shares to investors as soon as the week of Nov. 9, according to IPOScoop. The Wall Street Journal’s sources pointed to a listing in the week of Nov. 16, though they said that could change. Anthropic has not announced a date, and it said the offering will be made only by means of a prospectus once the registration takes effect.
