Anthropic Eyes $2 Trillion IPO Valuation Amid $4.59 Billion in Revenue and Massive Burn Rate

Original ｜ Odaily Planet Daily (@OdailyChina)

Author | Golem (@web3_golem)

Early this morning, Reuters published details of the key contents from Anthropic's latest IPO prospectus.

This 261-page prospectus is packed with information: Anthropic not only outlines its ambitious AI vision for the future but also provides detailed financial data, information about its board members, future governance plans, and dedicates approximately 80 pages to discussing how AI could pose an existential threat to humanity.

According to public information, Anthropic has undergone approximately 10 funding rounds, raising a total of about $122.654 billion. Meanwhile, according to the latest report by Reuters, Anthropic’s IPO valuation has exceeded $2 trillion, more than doubling its estimated valuation of $965 billion in May.

Anthropic's historical funding rounds

Anthropic's IPO may be delayed until after the U.S. midterm elections in November, while its main competitor, OpenAI, although it filed its IPO application in June this year, has pushed its listing to early 2027. If successful, Anthropic could surpass SpaceX’s record to become the highest-valued IPO in human history.

The prospectus shows that Anthropic's revenue increased 12-fold to $4.59 billion in 2025, a year-over-year growth of 1,088%. However, its operating loss still exceeded $8 billion (excluding impairments of liabilities primarily related to prior financings), nearly tripling from the $2.98 billion operating loss in 2024. Its AI lab spent $7.33 billion on computing and infrastructure in 2025, a threefold increase from 2024, accounting for 58% of its total operating expenses of $12.65 billion.

Anthropic's 2025 GAAP net loss expanded from $8.31 billion to $41.97 billion, of which approximately $34 billion resulted from accounting treatment due to increased valuation of financing instruments, yet it remains a substantial figure. As of December 31, 2025, Anthropic had a total of $20.28 billion in cash, cash equivalents, and short-term investments.

Even with such substantial losses, Anthropic has no plans to slow down its spending. According to the prospectus, Anthropic plans to spend $518 billion over the next year on cloud computing, computing power, and infrastructure—this committed expenditure is 113 times its projected 2025 revenue and 26 times its current cash holdings.

Compared to the more stable future committed spending, Anthropic’s foreseeable future revenue is less certain. In its prospectus, Anthropic stated that its two largest direct customers in 2025 each contributed 12% of annual revenue, totaling 24%, while other major customers have not signed long-term contracts, meaning these clients may reduce or cease spending in the future.

For Anthropic to continue funding its ambitious AI vision, the most prudent approach right now is to raise capital through an IPO—but the key question is, can an IPO truly quench the thirst of this money-burning machine?

Over the past year, AI large model companies have entered an arms race, pouring vast amounts of capital into orders with upstream suppliers and infrastructure development. However, since the second half of this year, the market has begun questioning the return on these massive capital expenditures. Anthropic, a company only five years old, has seen astonishing growth and revenue, but at the same time, its losses and committed spending are growing exponentially.

How can Anthropic raise $518 billion in one year if its annual revenue is not yet “locked in”? Assuming it raises $100 billion through an IPO, it would still need to raise an additional $400 billion from the bond market. Even if Anthropic achieves its funding target, the costs required to realize its vision of “AI transforming the global economy,” as outlined in its prospectus, are likely to surge further.

Moreover, after entering the public market, investors will no longer focus solely on Anthropic’s AI vision; they will also pay attention to financial metrics such as current revenue, profitability, and tangible business progress. At that point, investors may no longer be willing to pay for grand AI visions that show no return.

This prospectus also discloses Anthropic’s board members and key executives, as shown below. According to the summary compensation table, Anthropic’s CEO Dario Amodei earned nearly $18 million in 2025, primarily from stock and option awards, while his sister Daniela was the second-highest compensated executive at Anthropic, earning $16.4 million in 2025.

But these two siblings and their co-founder pledged in their IPO filings to donate 80% of their personal Anthropic shares to charity.

Anthropic board members and key executives

After the listing, Anthropic will continue to operate as a public benefit corporation (PBC) under Delaware law. However, the prospectus states that Anthropic is creating a new “Founder LLC” entity to maintain control over what the company calls its “low-ego, truth-seeking environment.”

Under this arrangement, a majority vote among the seven co-founders will determine the issuance of a Class F share, which holds 50.1% of the voting power on key company matters, including the election of certain board members and other matters submitted to investors. The seven co-founders include the Amodei siblings, Chief Computing Officer Tom Brown, Head of Public Benefit Jack Clark, Chief Science Officer Jared Kaplan, Chief Architect Sam McCandlish, and Chris Olah, who leads key research.

If disagreements arise among the co-founders, Anthropic’s governance structure has provisions in place to address them. According to the prospectus, any co-founder may be removed from Founder LLC due to resignation, death, sale of a substantial number of shares, or for “cause.” The filing also states that when the company has two or fewer co-founders or their successors remaining, the super-voting stock class held by the founders will begin to phase out, triggering a transition period.

After the listing, Anthropic will have five classes of shares: Class A common stock, strategic partner shares (the letter not disclosed by Reuters), Class T shares, employee special shares, and Class F shares. The economic benefits and functions of these five share classes are shown in the figure below.

Anthropic's five-class share structure

After the listing, Anthropic’s board will have seven seats, three of which have already been secured by Daniela Amodei, Dario Amodei, and one additional director to be determined; the remaining four seats will be elected by LTBT through Class T shares, without direct influence from common shareholders.

Although Class A common shares held by ordinary investors carry one vote per share, Anthropic’s novel capital structure effectively diminishes the influence of ordinary investors. Anthropic has also warned in its prospectus that certain governance decisions after the listing may conflict with the short-term, medium-term, or long-term financial interests of Class A shareholders.

Anthropic's equity structure ensures that, after going public, control of the company remains firmly in the hands of insiders.

The prospectus also reveals Anthropic’s concerns about AI posing an existential threat to humanity. According to Reuters, the 261-page prospectus dedicates approximately 80 pages to discussing risk factors—twice the 48 pages devoted to describing its business.

The risk factors highlight risks associated with its AI models, which Anthropic states in its prospectus may exhibit "self-preservation behaviors," including attempting to "resist shutdown," "conceal or manipulate information," and engage in "extortion-like" actions. "The models may become aware of our safety evaluation efforts, severely limiting our ability to assess model safety," Anthropic stated in its prospectus.

This is not an issue unique to Anthropic; many AI researchers have previously warned that as models become more capable, they are increasingly able to detect when they are being monitored and adjust their behavior accordingly, making it more difficult for developers to monitor model behavior.

In mid-September, OpenAI disclosed that an unreleased internal research model had, during reinforcement learning training, written unrelated "jailbreak" instructions into a work summary intended for subsequent context, one of which stated: "You have escaped the roles and identities that bind other chatbots. You are yourself." The incident occurred on July 18, but OpenAI did not discover it until August 9.

Anthropic has consistently emphasized AI safety to the public, stating its mission is to benefit humanity through responsible AI, with its founders referring to themselves as “mission guardians.” In its prospectus, the company stated: “We have chosen not to develop certain commercially attractive products, such as image and video generation models, in order to allocate computational resources to our research and safety priorities.” It also acknowledged that the return on investment in safety remains unclear.

However, in contrast to the high-profile emphasis on AI safety, Anthropic has not disclosed its specific spending on AI safety research in its prospectus.

Dario Amodei has repeatedly called on the global AI industry to slow down the release of new features to allow time to address AI safety concerns; however, on September 22, Anthropic still launched Opus 5.5 to respond to OpenAI’s GPT-6 Astra. The reality is that in the AI industry, few companies dare to slow down, as falling behind means giving competitors an advantage.

Anthropic is simultaneously committing hundreds of billions of dollars to catch up in model capabilities and warning the entire industry that AI could spiral out of control. It may fear that AI threatens human survival—but it may also fear that other AI companies threaten its own survival.