Anthropic launched the roadshow for its stock market debut on Tuesday, setting a price of $70 to $78 a share for 825 million shares in a sale that could raise $64 billion.

Key Points

  • Anthropic opened the roadshow for its initial public offering on Tuesday 10 November, offering 825 million Class A shares at $70 to $78 each.
  • The sale could raise between $57.75 billion and $64.35 billion, and a separate $26 billion private placement takes total fundraising toward $90 billion.
  • At the top of the range, the company behind the Claude family of AI models would be worth about $1.4 trillion on a fully diluted basis.
  • Once trading begins, the stock will carry the symbol “ANT” on the Nasdaq, with the first trades reported for around 19 November.
  • The same filing says Anthropic has decided it has met its own threshold for AI speeding up AI research, based on expectations for future models rather than current ones.
  • A day earlier, Anthropic withdrew pricing details it had sent to bankers after calling them inaccurate.

Anthropic launched the roadshow for its stock market debut on Tuesday, setting a price of $70 to $78 a share for 825 million shares in a sale that could raise $64 billion.

What happened

Anthropic, the company behind the Claude family of AI models, launched the investor roadshow for its initial public offering on Tuesday 10 November. The same day, it sent the US securities regulator an amended version of its registration paperwork — the first version to carry a price range. Anthropic has asked for its Class A shares to be admitted to the Nasdaq Global Select Market, where they would trade under the symbol “ANT”.

The offering covers 825 million Class A shares. Anthropic itself is selling 801,351,354 of them, and existing stockholders are selling a further 23,648,646. At $70 a share the sale would raise $57.75 billion; at $78 it would raise $64.35 billion. Goldman Sachs and Morgan Stanley are the lead underwriters, and the underwriters have a 30-day option to buy up to 123.75 million more shares at the IPO price.

Alongside the public sale, Anthropic has agreed a separate $26 billion private placement of non-voting Class N stock, priced at the same per-share price as the IPO and dependent on the offering going ahead. Seven companies are taking part: AMD, Broadcom, Google, Micron, Microsoft, Nvidia and TSMC. Google and Nvidia are each putting in $10 billion. Taken together, Anthropic estimates net proceeds of between $84.6 billion and $93.7 billion, the higher figure applying if the underwriters exercise their option in full.

The price range arrived a day after a false start. Late on Monday 9 November, Reuters reported that Anthropic had sent bankers a draft filing showing about 867 million shares at $70 to $80 each, worth up to $69.3 billion, and then emailed to say the pricing details were inaccurate and an updated filing would follow. The filing published on Tuesday offers fewer shares and tops out at $78. Anthropic declined to comment on the earlier draft.

The filing also sets out the scale of the business behind the listing. Revenue for the three months to 30 September 2026 was $17 billion, and revenue for the first nine months of the year was $33.2 billion. The same nine months produced a net loss of $104 billion, and the company reported computing and infrastructure commitments of $900.9 billion. It held nearly $132 billion in cash and short-term investments. Run-rate revenue reached $81 billion by the end of October, about $10 billion more than a month earlier.

Control will stay with the founders. As a public benefit corporation, Anthropic’s founders hold a Class F share carrying 50.1% of the voting power, and its Long-Term Benefit Trust will elect a majority of the board, so buyers of the Class A shares will have limited influence over the company. The filing also lists the shares the co-founders plan to sell: chief executive Dario Amodei, president Daniela Amodei, chief compute officer Tom Brown, Anthropic Institute head Jack Clark, chief science officer Jared Kaplan, chief architect Sam McCandlish and interpretability research lead Christopher Olah are each selling nearly 3.4 million shares, worth about $250 million each at the $74 midpoint of the range.

The same filing carries a disclosure that goes beyond the numbers. Anthropic said it has made a determination that it has met the threshold in its own Responsible Scaling Policy for “dramatically accelerated progress attributable to AI” — the point at which AI systems start speeding up AI research itself. The determination rests on internal estimates and expectations for future models, not on current ones, and the company said it has not designated any current model as exceeding the threshold. An earlier version of the filing, dated 2 November, had said only that future models may “imminently” meet it. A separate assessment published on 5 November by METR, an AI evaluation group given access to Anthropic’s internal data, estimated the company’s models are now improving about twice as fast as they did between March 2024 and October 2025, with most of the speed-up coming from AI doing AI research — a judgment METR itself called high-uncertainty.

Why it matters

The listing is the first time a frontier AI lab has asked the public markets to put a price on it, and the number is enormous. At the top of the range, Anthropic would be worth about $1.4 trillion on a fully diluted basis, according to CNBC and TechCrunch, while The New York Times put the range at $1.26 trillion to $1.41 trillion. That is well above the $965 billion valuation Anthropic reached in a May fundraising, but below the $2 trillion the company was reportedly targeting a couple of months ago — a comedown that mirrors the recent pullback in AI stocks, and one reason investors were already pushing back on the valuation before the range was set.

The deal also shows how the AI industry now finances itself. The seven companies buying into the private placement are the same companies that sell Anthropic its chips, run its cloud infrastructure and compete with it in AI — Google and Nvidia alone are committing $20 billion between them. That intertwining is part of the business model: Anthropic’s spending plans are so large that it is buying power plants and generators to run its data centres, and the filing’s $900.9 billion in computing commitments dwarfs its $33.2 billion of nine-month revenue. The IPO is, in effect, a public test of whether investors believe that spending turns into profit.

For ordinary investors, the offering is unusually open. TechCrunch reported, citing CNBC, that Anthropic expects up to 20% of the offering to go to retail investors, a far larger share than most large listings set aside. But the governance terms mean those buyers get almost no say: the founders’ voting control and the trust’s board majority are locked in from day one. The company also uses the terms “super intelligence” and “SI” through the filing, echoing a September executive order from President Donald Trump that directed government departments to use the new terminology for AI — a small but deliberate signal about how the company wants to be seen in Washington.

The research threshold disclosure matters for a different reason. Anthropic’s own policy requires it to publish further reports once the threshold is met, and it says it is weighing changes to the policy that would require human review and approval of certain research decisions. The company has said it does not expect to let Claude design and build successive frontier models without meaningful human supervision. Investors buying into the IPO are therefore buying into a company that has just told regulators its own technology may be about to accelerate its own development — and that it intends to keep humans in charge of that process.

What happens next

The roadshow runs this week, and the final price will be set after it ends. Reuters reported that pricing is expected in the week of 16 November, while CNBC reported that Anthropic plans to price on 18 November and start trading on 19 November, and The Wall Street Journal reported trading is planned for 19 or 20 November. None of those dates comes from Anthropic itself, and the filing leaves the delivery date blank. The regulator has received the paperwork but has not yet declared it effective, and until it does, no shares can be sold and no purchase orders accepted.

Two things will decide how the listing is judged. The first is the price itself: it can land inside the $70 to $78 range, above it or below it, and the range has already been trimmed once, from the $70 to $80 sent to bankers on Monday. The second is the first day of trading, when the market will say whether a $1.4 trillion valuation for a company that lost $104 billion in nine months is a fair price for the AI boom or a sign of how far expectations have run ahead of the business. Anthropic’s latest models, Muse Ember and Claude Fable 5.2, launched the same week the roadshow began — a reminder that the company is still shipping products while it asks the market to value them.

Sources: Anthropic (company announcement and Form S-1 Amendment No. 2, US Securities and Exchange Commission), Reuters, TechCrunch, CNBC, The New York Times, The Wall Street Journal, dpa-AFX, METR. Published: 11 November 2026, Kuala Lumpur time.