Anthropic has launched its initial public offering roadshow, offering 825 million shares at a price range of $70 to $78 each. The pitch to institutional investors centers on one of the most closely watched listings in the AI industry. Pricing is expected on November 18, with trading on Nasdaq under the ticker ANT the following day, though those dates are based on people familiar with the deal and are not final until the company announces them.
The Offering Terms
Anthropic, a public benefit corporation, began its roadshow on November 10, 2026. At the top of the range, the base offering would raise about $64.35 billion before fees. Underwriters also have an option to buy nearly 124 million additional shares within 30 days, worth roughly $965 million at the top price.
About 23 million of the shares will come from the company's seven co-founders, according to the offering details reported by SiliconANGLE, and each is set to receive $250 million. Those founder sales are a notable detail because they signal how much of the deal provides liquidity to insiders rather than fresh capital for the business.
The Valuation Debate
At the top of the range, Anthropic would be valued at about $1.405 trillion on a fully diluted basis, according to CNBC. That compares with a $965 billion valuation from the company's May funding round. The gap is large, and it reflects how much the price has moved as the listing approached.
CNBC reported that some investors had expected a lower figure, with one calling $1.5 trillion a stretch. Anthropic had previously been aiming for a valuation of about $2 trillion, so the current range represents a meaningful reset. Public-market investors will decide whether the business can support the price, and the number that matters most is what the shares trade at once they begin to change hands.
Who Is Buying
Anthropic is also selling $26 billion of non-voting shares in a separate private placement. Seven major suppliers and partners, including Nvidia, AMD, Broadcom, Google, Micron, Microsoft and TSMC, will buy that stock at the IPO price. Those are the companies that build the chips, memory and cloud infrastructure that frontier models depend on, so their participation ties the listing to the broader AI supply chain.
According to people familiar with the deal cited by CNBC, the book is already covered at the price range on an allocable basis. CNBC also reported that Anthropic is targeting about 20 percent of the shares for retail investors, well above the typical 5 to 10 percent. A large retail allocation widens who owns the company, but it can also make the stock more volatile in the first weeks of trading.
What the Structure Signals
Taken together, the terms show a company trying to raise a very large amount of capital while keeping strategic partners close. The private placement at the IPO price gives suppliers a stake in the company, and the public offering gives ordinary investors a way in. Both sides have an interest in the company's compute deals succeeding, which makes the listing as much a statement about infrastructure as about the software itself.
The founder sales also deserve attention. Co-founders selling a portion of their holdings at the offering is common in large listings, but the amount will be scrutinized by investors who want to see long-term commitment. How the company explains those sales in its public filings will shape early sentiment.
Why It Matters
Anthropic's listing will set a reference point for how public markets value frontier AI companies. A strong debut could encourage other AI labs to pursue listings, while a weak one would give skeptics an easy example. The range also tests whether investors accept valuations that are far above the company's last private round.
The risks are real. Pricing is not final, and the company still must announce its terms. Investors will weigh the revenue growth that justifies the valuation against the heavy spending needed to run frontier models. For now, the roadshow shows how much the AI trade depends on whether public investors are willing to pay for the next wave of compute-heavy companies.
