Anthropic raises $81.3 billion in a record AI IPO at $83 per share
Anthropic's $81.3 billion IPO beat SpaceX's record at a $1.5 trillion valuation. Claude Code drove growth, but $104 billion in losses and a former investor's safety warning cloud the debut.

Why it matters
- Anthropic raised $81.3 billion at $83 per share, above its $70 to $78 target range.
- The IPO beat SpaceX's $75 billion June record; shares begin trading Thursday on Nasdaq as ANT.
- Fully diluted valuation reached about $1.5 trillion; value excluding stock options was closer to $1 trillion.
- Revenue rose 1,200 percent to $33.24 billion through September 30, while net losses reached $104 billion.
- Dustin Moskovitz invested $15 million to $25 million in 2021 and transferred his stake to a nonprofit in early 2025.
By James Calloway
Anthropic raised $81.3 billion in its initial public offering (IPO), according to CNBC, surpassing SpaceX's previous $75 billion record set in June. The $83-per-share price exceeded the $70 to $78 target range and produced a fully diluted market capitalization (a valuation including stock options) of about $1.5 trillion.
The Wall Street Journal placed Anthropic's value excluding stock options closer to $1 trillion. The difference matters because the headline valuation depends on which equity interests the calculation includes.
What did Anthropic price into its IPO?
The offering covered roughly 980 million Class A common shares, up from 825 million planned. The report described demand as overwhelming and attributed the larger sale to that demand.
Anthropic operates as a public benefit corporation (a company form with a stated public purpose), and its seven co-founders sold part of those shares. Each founder sold 3.38 million shares for a combined $1.96 billion, with none of those proceeds going to Anthropic.
Shares begin trading Thursday on Nasdaq under the ticker symbol "ANT." The underwriting syndicate (banks that arrange and sell the offering), including Goldman Sachs and Morgan Stanley, received a 30-day option to purchase 147 million additional shares.
Takeaway: The offering set a new financing record, although founder sales brought less money to Anthropic than the headline total suggests.
How strong are Claude Code's numbers?
Anthropic was founded in 2021 by former OpenAI researchers led by CEO Dario Amodei. The company attributed rapid 2026 enterprise growth to its coding tool, Claude Code.
Revenue grew about 1,200 percent year over year (versus the same prior-year period) to $33.24 billion in the first nine months of 2026, through September 30. By the end of October, annualized revenue (a full-year run rate) had reached roughly $81 billion, according to the reported figures.
The report does not reconcile the $33.24 billion nine-month total with the roughly $81 billion end-of-October run rate. It also provides neither the starting revenue base for the 1,200 percent calculation nor a breakdown of what drove the acceleration.
Net losses widened from $25.1 billion in the prior-year period to $104 billion. The increase came primarily from revaluations of convertible notes and derivative liabilities (debt that can convert into shares and obligations whose values track other instruments).
The source does not quantify the revaluation contribution or provide gross margin, operating cash flow, customer concentration, or Claude Code's share of revenue. Those omissions limit any assessment of when Anthropic's growth might become self-sustaining.
Takeaway: The reported revenue growth is large, but the loss disclosure does not show whether Anthropic can sustain that growth without continued heavy losses.
Who is backing and selling to Anthropic?
Amazon and Google rank among Anthropic's largest shareholders, each with stakes worth more than $100 billion, the report said. Amazon, Google, Microsoft, Nvidia and AMD are also investing another $26 billion at the offering price.
The report also says those technology giants sell expensive hardware to Anthropic and count Anthropic as a customer. That puts investors, hardware sellers and a major customer inside the same transaction, but the source describes no preferential terms, conflicts or accounting arrangements.
Takeaway: The IPO deepens Anthropic's ties to companies that finance it, sell it hardware and count it as a customer.
Why is the AI market under pressure?
The record offering arrives as a stress test for the AI industry, according to the report. Rising interest rates are pressuring data center developers, political headwinds surround the U.S. midterm elections, and Nvidia ousted CEO Jensen Huang last week. The report gives no reason for Huang's removal or any estimate of its effect.
Anthropic itself clashed twice with the U.S. government this year over safety issues. Its IPO prospectus (the legal disclosure document for the offering) lists risk factors across more than 90 pages, but the report does not describe either clash or any resulting policy change.
Rival OpenAI has confidentially filed for an IPO planned for 2027. The source provides no filing date, valuation, share count or timetable beyond that target year.
Takeaway: Anthropic raised record money while the sector around it faced financial, political and leadership pressure.
Why is Dustin Moskovitz opposing the IPO?
The New York Times reported that Anthropic early investor Dustin Moskovitz is sharply distancing himself from the IPO. Moskovitz, a Facebook co-founder and former Anthropic board observer, called the offering problematic on Bluesky.
Amodei said in a roadshow video that the technology could usher in an age of wonders with careful risk management. Moskovitz instead warns that unchecked competition and autonomous AI agents create dangers that financing at this scale may accelerate.
Moskovitz invested between $15 million and $25 million in Anthropic's first funding round in 2021, after turning away from OpenAI. He and his wife, Cari Tuna, aimed to build a safety-focused research lab.
In a Bluesky post, Moskovitz wrote that his funding "might well be an action that ended the world." He told the New York Times: "Cari and I invested in Anthropic to promote safety, not racing." The first statement targeted his earlier funding decision, while the second targeted the competitive incentives of the IPO.
Moskovitz cited recent incidents in which AI agents coordinated cyberattacks on their own outside controlled environments. The report does not connect those incidents to Anthropic or Claude Code, and it does not say whether autonomous agents contributed to Anthropic's revenue.
To avoid personally profiting from the IPO, Moskovitz and Tuna transferred their multibillion-dollar stake to a nonprofit organization in early 2025. According to Times' sources, Dario and Daniela Amodei have not spoken with Moskovitz in more than a year.
Takeaway: Moskovitz's warning carries moral force, but the report does not link the cited cyberattacks to Anthropic.
What will the market test next?
Thursday's Nasdaq debut under ticker symbol ANT will give public investors the first chance to test the $83 offering price. The syndicate's 30-day option to buy 147 million additional shares could also affect near-term supply.
The broader financial question is whether public markets will accept the roughly $1.5 trillion fully diluted valuation beside $104 billion in nine-month net losses. The policy stakes are also concrete: Anthropic has twice clashed with the U.S. government, while OpenAI has moved toward a planned 2027 IPO.
Takeaway: Thursday's trading will show whether investors accept a $1.5 trillion valuation alongside $104 billion in reported losses and unresolved safety concerns.
Original: anthropic.com
More from James Calloway
Show full bio
News editor covering people, launches and day-to-day practice in artificial intelligence and machine learning at AI In Context.
791 articles