November 10, 2026, (Inside AI) — Anthropic's initial public offering is set to raise as much as $64.4 billion, according to the latest draft prospectus seen by sources on Tuesday. The AI company and some of its investors plan to sell shares in an indicative price band of between $70 and $78 per share. Deal details are expected to be announced soon. Anthropic did not respond immediately to a request for comment.
The offering would rank among the largest technology IPOs in history. At the top of the range, Anthropic's raise would exceed the $25.9 billion raised by Saudi Aramco in 2019, the world's largest IPO to date, according to data from Dealogic. The filing arrives as investors race to gain exposure to the companies building the infrastructure and models behind the generative AI boom.
Anthropic, founded in 2021 by former OpenAI researchers including Dario Amodei and Daniela Amodei, has positioned itself as a safety-focused AI developer. Its flagship model family, Claude, competes directly with OpenAI's GPT series and Google's Gemini. The company has raised more than $15 billion in private funding, with backers including Google and Amazon. A public listing would give those investors a path to liquidity and provide Anthropic with a war chest for compute and talent.
The IPO also tests public market appetite for a company that has yet to report sustained profitability. Anthropic's revenue has grown quickly, but its training costs remain enormous. The prospectus is likely to face scrutiny over customer concentration and the long-term economics of foundation models.
Why The Price Band Matters
The $70 to $78 range implies a valuation that could approach $300 billion on a fully diluted basis, according to calculations based on the draft. That would make Anthropic one of the most valuable companies to ever list in the US. The band is narrow, a signal that underwriters have gauged strong demand from institutional investors. A narrow range also reduces the risk of a first-day pop that leaves money on the table for the company.
Some analysts question whether the valuation is justified. Sarah Kline, a senior research analyst at Forrester, said the market is pricing in years of hypergrowth. "The IPO market has not seen a company with this combination of revenue scale and losses since the early days of cloud computing," Kline said. "Investors are being asked to underwrite a future where Anthropic captures a large share of enterprise AI spending."
The offering also comes amid a broader wave of AI listings. Rival OpenAI has been reported to be exploring a public offering, though no formal filing has emerged. Other AI infrastructure firms, including chipmakers and data center operators, have seen their valuations surge. The Anthropic IPO could serve as a bellwether for the sector.
Commentary: The decision to price Anthropic at a potential $300 billion valuation rests on an assumption that enterprise AI spending will keep compounding at current rates. But the company's own prospectus will need to show a credible path to margins that justify that multiple. Unlike software firms of the past, Anthropic must pay for massive compute bills that scale with usage. If public investors apply traditional profitability tests, the stock could face pressure after the lock-up period expires. The IPO's success may hinge less on the AI narrative and more on unit economics that have yet to be proven at scale.
Anthropic's offering is being managed by a syndicate of banks, though the draft prospectus did not name them. The company plans to use proceeds for general corporate purposes, including research and development and potential acquisitions. The roadshow is expected to begin in the coming weeks, with pricing likely before the end of the year.
The IPO would also create a new class of public shareholders with a say in Anthropic's governance. The company has structured its board to include safety experts, a feature that may draw both support and criticism from investors focused on returns. How Anthropic balances its public market obligations with its safety mission will be closely watched.
Shamil has a low tolerance for AI that only works in controlled demos. He covers agentic systems and generative AI with the skepticism of someone who has seen too many “revolutionary” tools quietly disappear six months after launch.
