Anthropic aims for a valuation of up to $1.4 trillion in a potential stock market listing.

Anthropic is preparing for an initial public offering (IPO) with a target valuation of $1.4 trillion (£1.1 trillion)—a figure lower than the $2 trillion previously anticipated, before investors began to question the sustainability of the artificial intelligence boom.
The developer of the Claude chatbot had reportedly planned for a $2 trillion IPO on Wall Street; however, a U.S. regulatory filing submitted on Tuesday indicates that this target has been revised downward.
Anthropic intends to offer 825 million shares in the IPO at a maximum price of $78 per share, aiming to raise up to $64.4 billion. The California-based company also plans to secure an additional $26 billion through a separate private placement with cornerstone investors, including Google parent company Alphabet and chipmaker Nvidia.
Should investor demand prove strong, Anthropic could raise an extra $9.7 billion via a “greenshoe option” (over-allotment provision). If the shares are sold at the top of the price range, the total amount raised by Anthropic would reach approximately $100 billion.
If the share price reaches the upper end of the $70–$78 range, Anthropic’s fully diluted market capitalization would hit $1.4 trillion—a figure that accounts for all of the company’s existing common shares, including employee stock options.
Shares of Anthropic are expected to begin trading next week. The outcome of last week’s U.S. midterm elections—which saw victories for opponents of data centers—has intensified investor concerns regarding the artificial intelligence boom. This has further fueled existing doubts about whether corporate investments in AI infrastructure will yield quick returns, or indeed any returns at all.
Brian Hopkins, a principal analyst at Forrester Research, noted that the midterm elections provided investors with an opportunity to recalibrate valuations—a process already underway due to concerns over the financial sustainability of AI-related infrastructure development. According to Hopkins, the election results have only exacerbated these anxieties.
“The market needed a catalyst to adjust valuations downward,” he said, “and the jitters surrounding the U.S. elections provided exactly that.”
He added that, as a result, the valuation for Anthropic has also been revised downward from the previously anticipated $2 trillion.
The largest stock market debut in history was that of Elon Musk’s SpaceX, which entered the U.S. market with a valuation of $1.78 trillion, raising $75 billion. SpaceX—whose assets include Musk’s rocket launch business and the social media platform X—currently holds a market capitalization just shy of $2 trillion.
However, the $100 billion that Anthropic could potentially raise in this offering would exceed the amount raised by SpaceX.
Anthropic CEO Dario Amodei plans to sell approximately $260 million worth of shares in the IPO, with his six co-founders also set to sell portions of their holdings. Based on the upper end of the offering price range, the value of each co-founder’s stake would exceed $18 billion.
A Chinese espionage scandal involving Nvidia—a company central to the AI investment landscape—also threatens to cast a shadow over the entire industry; thus far, however, the share prices of other companies remain unaffected.
According to the latest documents released by Anthropic, the company’s annual revenue stood at $81 billion as of October 31—an increase of $10 billion from the previous month.
A significant portion of Anthropic’s expenditure is directed toward computing power for its data centers, which is essential for the development and operation of its AI systems. Its prospectus reveals that the company’s commitments for computing power procurement and lease agreements total over $70 billion for the next 12 months, with that figure projected to exceed $450 billion by 2030, followed by an additional $450 billion in expenditures through 2045.
In a podcast appearance earlier this year, Amodei explained that investments in computing power must be balanced with revenue growth.
“I could go out and buy $1 trillion worth of compute by the end of 2027. And if I don’t have $1 trillion in revenue—even if I have $800 billion—there is no force on Earth, no hedge, that could prevent me from going bankrupt,” he told podcast host Dwarkesh Patel.
On Tuesday, the Financial Times reported that Goldman Sachs—which is leading the IPO alongside Morgan Stanley—projects that Anthropic’s annual revenue will reach $600 billion by 2030.