Anthropic IPO: Key Numbers From the Claude Maker's S-1

Anthropic, the San Francisco AI company that makes the Claude family of models, is officially on the path to becoming a public company. On Monday, Oct. 26, 2026, it announced that it had filed a registration statement on Form S-1 with the U.S. Securities and Exchange Commission for a proposed initial public offering (IPO) of its Class A common stock. The filing on SEC EDGAR gives the public its first full look at the company's finances. Below are the numbers and details that matter most, in plain English.

Not investment advice. This article explains what the public filing says. It is not a recommendation to buy or sell any security.

What we know about the Anthropic IPO

• Ticker: ANT, on the Nasdaq Global Select Market.
• Price and size: Not yet set. Anthropic said the number of shares and the price range "have not yet been determined."
• Banks: Goldman Sachs and Morgan Stanley are the lead underwriters, with about 20 other banks in the syndicate, CNBC reported.
• Status: The registration statement "has not yet become effective," so no shares can be sold yet.
• Background: Anthropic first filed confidentially with the SEC in June, which let regulators review its financials before they were made public, according to CNBC.

What is an S-1? It is the registration statement a company files with the SEC before selling stock to the public for the first time. It contains audited financial statements, a description of the business and a long list of risk factors. An IPO is the first sale of a company's stock to the public, as the SEC's Investor.gov glossary explains.

The key numbers in Anthropic's S-1

Sources: CNBC's S-1 coverage and the S-1 filing.

Why the loss is so large

The headline $80.7 billion loss looks alarming, but context matters. CNBC reported that roughly $72 billion of it comes from accounting charges for changes in the value of convertible notes, which rose as Anthropic's valuation increased. The Verge, citing earlier Wall Street Journal reporting, also noted that most of the loss is accounting-related.

The operating picture is different. CNBC calculated that cost of revenue was $2.38 billion against $16.29 billion of revenue in the first half, which implies a gross margin of about 85%, up from about 12% a year earlier. The filing warns margins may swing as large amounts of new computing capacity come online.

Where the revenue comes from

• Consumption-based revenue (customers paying for usage) was $13.1 billion, or 80% of the total, in the first half.
• Subscriptions were $3.2 billion, or 20%.
• Cloud marketplaces (Amazon Web Services, Google Cloud and Microsoft Azure) accounted for 48% of revenue in the first half of 2026.
• International revenue was $4.5 billion, about 27.8% of the total.

All four figures are from CNBC's reading of the filing. The filing also says no single customer made up more than 10% of revenue in the first nine months of 2026.

The giant compute bill

Training and running AI models needs huge amounts of computing power, and the S-1 shows how much Anthropic has committed to buy:

• SpaceX: about $3.1 billion a month for Nvidia GPU capacity once fully delivered, up to $120.5 billion through 2029.
• Google: at least $114.5 billion for TPU chips through 2033.
• Amazon: $110 billion for cloud services and Trainium chips through 2036.
• Microsoft: at least $44.9 billion between November 2026 and May 2033.

These commitments were reported by The Verge and CNBC. The company warns that if it cannot secure enough compute, it risks "falling behind the technological frontier."

Who will control Anthropic after the IPO

Anthropic is a public benefit corporation (PBC): a for-profit company that must, under Delaware law, balance shareholder returns with a stated public benefit. For Anthropic, that purpose is "the responsible development and maintenance of advanced AI for the long-term benefit of humanity," according to the prospectus.

Its share structure is unusual:

• Class A: the shares sold to the public, with one vote each.
• Class F: a single share held by a "Founder LLC" of the seven co-founders, including CEO Dario Amodei and President Daniela Amodei. It carries 50.1% of total voting power on most shareholder votes.
• Class T: held by the Long-Term Benefit Trust, which elects directors who make up a majority of the board.

Anthropic itself says this structure "has no direct public company precedent" and could make it harder to replace directors or for shareholders to receive a takeover premium, CNBC reported.

Risks the filing highlights

The S-1 includes more than 90 pages of risk factors. Among them:

• Competition from OpenAI and Google, the only two competitors named.
• Government disputes, including a Pentagon "supply chain risk" designation that is still being litigated.
• Supply chain dependence on TSMC in Taiwan for advanced chips.
• Data center moratoriums being considered in some U.S. states.
• AI safety and misuse, including the risk that safeguards fail.

What's next

1. Price range. Anthropic will amend its filing to set the number of shares and a price range.
2. Roadshow. Executives pitch the offering to investors. The Information reported the roadshow could start in about two weeks, with a listing targeted for the week of Nov. 16. Anthropic has not confirmed a date.
3. SEC effectiveness and pricing. Only after the SEC declares the registration effective can the IPO be priced and shares begin trading under ANT.

The timing comes as investors debate AI spending. For more on the money side of the trends Americans are following, see our Money section and Tech coverage, and catch up on the 2026 midterm elections.

The bottom line

Anthropic's S-1 shows a business whose revenue has grown roughly fourteenfold in a year, a very large accounting-driven loss, hundreds of billions of dollars in computing commitments and a governance structure that keeps control with its founders and a safety-focused trust. The final price, size and date are still unknown.