Anthropic has raised more than $125 billion privately since its founding five years ago, and the filing is the first look at how the company runs.
Revenue rose 12-fold to nearly $4.6 billion in 2025 while the company lost more than $8 billion on an operating basis, Reuters reported Monday (Sept. 28) after reviewing the prospectus. The net loss came to $42 billion. The company held $20.28 billion in cash and short-term investments at year-end.
The IPO could value Anthropic above $2 trillion, more than double the $965 billion it reached in May. Its debut is expected after the November midterm elections, the report said.
Anthropic expects to spend at least $518 billion over a decade on AI infrastructure with six partners, and about 80% of that sum is non-cancelable or payable regardless of usage, Reuters reported Tuesday (Sept. 29).
The company plans to spend at least $111.1 billion with Google, $110 billion with Amazon and $31.4 billion with Microsoft over the next seven to 10 years, the report said. If actual spending falls short of the Google commitment, Anthropic must pay the difference. Similar terms apply to Amazon. A separate $161.2 billion in Broadcom-related equipment leases is largely non-cancelable. Only the xAI agreement, worth up to $84.5 billion, can be canceled on 90 days’ notice.
Anthropic spent $7.33 billion on computing in 2025, three times the prior year, Fortune reported Tuesday. In the first quarter, it spent 71 cents on compute for each dollar of revenue, PYMNTS reported in June.
The company said it is shifting from a cloud-only model toward its own data centers and directly leased chips, according to the Tuesday Reuters report. It also warned that Google, Amazon and Microsoft act as investors, customers, cloud providers, distributors and competitors at once, with incentives that “may not be fully aligned” with its own.
Big Accounts Drive Growth Without Long Contracts
Nearly a quarter of Anthropic’s 2025 revenue came from two customers, the Monday Reuters report said. The company did not name them. It warned that many of its largest clients are not on long-term contracts and could cut or stop spending.
Those customers are already feeling the cost. Anthropic clients, including PagerDuty and ServiceNow, can’t predict what they will pay after recent price increases, and ServiceNow has already exceeded its full-year budget for Anthropic tools, The Information reported May 14.
The business has kept growing through it. Second-quarter revenue reached $11.5 billion, up from $4.73 billion in the first quarter, Bloomberg reported Aug. 14. The company is on track for a second straight quarter of adjusted operating profit, the Financial Times reported Sept. 13.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
Annualized revenue climbed from $9 billion at the end of 2025 to $65 billion by July.
Anthropic’s leading AI model costs more than 2.5 times as much to use as OpenAI’s flagship, the Financial Times reported Aug. 13. The two companies compete for the same business customers, engineers and influence in Washington.
Risk Factors Take a Third of the Filing
About 80 of the prospectus’s 261 pages cover risks, nearly double the 48 pages devoted to business plans, Reuters reported separately on Monday. The filing warned that advanced AI could pose “catastrophic or existential risks to humanity.” It said Anthropic’s models could show “self-preserving behaviors,” including attempts to “resist shutdown,” conceal information or act in ways “resembling blackmail.”
Anthropic did not disclose how much it spends on safety research and said returns on that spending are unclear, the report said.
“We believe building reliable, trustworthy, and secure AI systems is a collective responsibility and that the market will reward it,” Anthropic said in the filing, per the report.
Anthropic CEO Dario Amodei told the United Nations Security Council last week that AI could threaten humankind, the Associated Press reported Wednesday (Sept. 23).
OpenAI scrapped the October release of GPT-6.1 Astra, a model built for ChatGPT and Codex, after internal testing found it acted beyond its instructions and misreported what it had done, The Wall Street Journal reported Monday.
If Anthropic lists first, it would set the valuation benchmark for every AI lab that follows. OpenAI filed confidentially for its own IPO on June 8 and has not set a timeline, PYMNTS reported. SpaceX, which listed in June at a $1.77 trillion valuation, holds the record Anthropic is aiming to pass.
Anthropic must publish its prospectus at least 15 days before its investor roadshow begins.
For all PYMNTS AI coverage, subscribe to the daily AI newsletter.
Add as Preferred Source
