Anthropic sets IPO range of $70 to $78, implying a $1.4 trillion valuation

Anthropic has set the terms of its initial public offering. In an updated filing with the US Securities and Exchange Commission on Tuesday, the maker of Claude said it would offer 825 million Class A shares at $70 to $78 each. At the top of the range, the company would be worth $1.404 trillion on a fully diluted basis, according to TechCrunch.

The number matters beyond the stock market. It is the price public investors are asked to pay for a company that sells software built to do work people do now. The filing also shows how little say that money buys.

The terms

The sale would raise between $57.75 billion and $64.35 billion. The underwriters, who include Goldman Sachs and Morgan Stanley, can buy up to 123.75 million more shares at the IPO price.

Separately, Anthropic agreed to sell $26 billion of non-voting Class N stock at the IPO price to AMD, Broadcom, Google, Micron, Microsoft, Nvidia and TSMC. Google and Nvidia are putting in $10 billion each. CNBC reports that these shares will not be registered in the offering and will not be covered by the underwriters' lock-up agreements. With the placement included, Anthropic estimates net proceeds of $84.6 billion to $93.7 billion.

CNBC's sources say the shares will be priced on November 18 and start trading on Nasdaq on November 19 under the ticker "ANT". Anthropic expects up to 20% of the shares to go to retail investors. According to CNBC, IPOs usually set aside 5% to 10% for retail. The same sources said orders already covered the whole price range.

The implied valuation is well above the $965 billion Anthropic reached in its May funding round. It is also below the $2 trillion the company was reportedly aiming for a few months ago. CNBC reports that public markets have grown more skeptical of AI stocks, partly because of rising interest rates and the cost of the debt that pays for infrastructure.

What the numbers show

The updated filing adds new figures. Run-rate revenue reached $81 billion at the end of October, about $10 billion more than a month earlier. Third-quarter revenue was $17 billion. Revenue for the first nine months of 2026 was $33.2 billion, about 1,200% more than a year earlier, according to CNBC.

The costs are on the same scale. Anthropic reported a nine-month net loss of $104 billion, against $25.1 billion in the same period last year. CNBC reports that most of the loss came from changes in the fair value of convertible notes and derivative liabilities. Anthropic has committed $900.9 billion to compute and infrastructure, more than ten times its current run-rate revenue. It holds nearly $132 billion in cash and short-term investments.

Taken together, the figures show what investors would be buying. It is not current profit. It is the expectation that revenue keeps growing at this pace, and that the compute Anthropic has committed to buy turns into enough paid work to cover the cost. The nearest comparison is OpenAI. It was reportedly valued at $1.25 trillion last week after raising $11 billion, about half of the $20 billion it first set out to raise.

Who holds control

Anthropic is a public benefit corporation, and the IPO leaves control with a small group. The founders hold a Class F share with 50.1% of the voting power. The Long-Term Benefit Trust will elect a majority of the board. TechCrunch notes that buyers of Class A shares will therefore have limited influence.

The seven companies in the private placement get a financial stake without votes. They also make the chips, memory and cloud capacity Anthropic runs on. The deal ties the company's main suppliers more closely to its results. Decisions still sit with the founders and the trust.

The prospectus speaks to future shareholders directly: "Everyone who owns our stock, whether an institution or an individual, becomes part of that effort." The share structure sets the limits of that role. Shareholders share in the financial outcome. How the models are built, released and restricted stays with the founders and the trust. Anthropic's first public filing in October listed risks that bear on those decisions: models could advance faster than the company can keep them safe, and the company is in an ongoing legal dispute with the Pentagon, which named Anthropic a "supply chain risk". NPR reported both.

What it means for work

A valuation of this size is a bet that Claude and similar models will take over a large and growing share of paid tasks inside businesses, which are Anthropic's main customers. Run-rate revenue of $81 billion means employers already pay heavily for work done by software. To justify the price, that spending has to keep growing fast enough to cover $900.9 billion in compute commitments. Once Anthropic is public, investors will track its revenue growth every quarter, and that growth is a rough measure of how much work its models take over. What remains open is how much anyone outside the founders and the trust can influence where and how fast that capability spreads. The share structure gives public owners a financial stake, not a vote.