←News detail

The PaperUpdated Oct 27 13:26

Anthropic files a 335-page prospectus: first-half revenue of $16.286 billion, net loss of $80.748 billion

AI summary

On October 26, Anthropic submitted an S-1 registration statement to the SEC, planning to list on the Nasdaq Global Select Market under the ticker symbol ANT; the number of shares to be issued, the price range and the date have not yet been disclosed. Revenue in the first half of 2026 was $16.286 billion, up about 13-fold year over year, with a net loss of $80.748 billion, of which more than $72 billion came from changes in the valuation of convertible notes; usage-based billing revenue was $13.1 billion, accounting for 80% of total revenue. The company also disclosed more than $430 billion in cloud computing agreement commitments over the coming years, and warned that its models' ability to evade monitoring mechanisms is increasing.

Why read this

SelectedAI score 89

The prospectus discloses Anthropic's revenue structure, computing commitments and governance structure, which can be used to observe the funding and risk profile of leading model companies.

Topics

Tags

Body · AI translation

Download the app

+1Anthropic's prospectus revealed: first-half revenue surged 13-fold, net loss exceeded $80 billion, warning of AI overreach risks

The Paper reporter Hu Hanyan 2026-10-27 07:16 Source: The Paper ∙ 10% Company > Listen to full text Font size

AI (artificial intelligence) startup giant Anthropic has filed a 335-page prospectus, disclosing the company's explosive growth and the risk factors behind it.

On October 26 local time, Anthropic submitted an S-1 registration statement to the U.S. Securities and Exchange Commission (SEC), taking a key step toward one of the largest initial public offerings (IPOs) in history. The company plans to issue Class A common stock and has applied to list on the Nasdaq Global Select Market under the ticker symbol "ANT." However, the filing has not yet disclosed the number of shares to be issued, the price range, or the specific offering date.

Goldman Sachs is listed as the lead underwriter for Anthropic's IPO, followed by Morgan Stanley, with other investment banks including JPMorgan Chase, Citigroup, Barclays, Wells Fargo, Bank of America, Deutsche Bank, RBC Capital Markets, UBS Investment Bank and others. In the coming weeks, these institutions will engage with investors to determine Anthropic's listing valuation.

First-half revenue grew about 13-fold, massive computing commitments highlight funding needs

In the prospectus, Anthropic divides its business into three parts: models, applications, and agents.

In the first half of 2026, the company's revenue reached $16.286 billion, an increase of about 13-fold from $1.163 billion in the same period last year, with a net loss of $80.748 billion; full-year 2025 revenue was $4.585 billion, with a net loss of $41.965 billion.

In the first half of this year, the company's operating loss narrowed to $693 million from $2.135 billion in the same period last year, while adjusted operating profit swung from a loss of $1.8 billion to a profit of $1.6 billion. During the same period, the company's net cash flow from operating activities reached $3.9 billion, compared with a net outflow of $1.5 billion in the same period last year.

Anthropic said revenue growth and improved computing resource and operational efficiency drove the improvement in the company's operating performance. Of the massive net loss in the first half of this year, more than $72 billion was related to valuation changes in convertible notes, due to the company's rising valuation.

The remaining losses mainly came from nearly $9 billion in computing and infrastructure spending, compared with $7.3 billion in related spending for all of 2025.

According to the company, as of September 30, among Anthropic's enterprise customer base, more than 11,900 customers had contributed revenue of $100,000 or more over the trailing 12 months (TTM). Among them, more than 1,900 customers contributed revenue of $1 million or more, and about 250 customers contributed $10 million or more.

The prospectus also shows that usage-based billing is Anthropic's main source of revenue. In the first half of this year, Anthropic's usage-based billing revenue was $13.1 billion, accounting for 80% of total revenue; subscription revenue was $3.2 billion, accounting for 20% of total revenue. The company expects that for the foreseeable future, usage-based billing revenue will continue to account for the vast majority of total revenue.

It is worth noting that behind the rapid business expansion are heavy long-term spending commitments. According to calculations, Anthropic plans to significantly increase spending in the coming years, spending more than $430 billion to sign locked-in cloud computing agreements with cloud giants such as Amazon and Google. Including cancellable smaller agreements, the total will be as high as $637 billion.

Anthropic pointed out that its revenue faces risks from customer concentration, namely that a small number of high-value customers contribute a considerable proportion of revenue. In the nine months ended September this year, no single customer contributed more than 10% of revenue, but about a quarter of revenue in the same period came from the top five customers.

The company added: "For the foreseeable future, we expect that we will continue to derive a substantial portion of our revenue from a limited number of customers, and in some cases, the proportion of revenue contributed by certain customers may increase in the future."

Corporate governance structure is also a noteworthy part of this IPO. Anthropic emphasized that the company has adopted a novel capital and corporate governance structure that "has no direct precedent among public companies": "Our governance structure may make it more difficult to replace directors and management, and may cause shareholders to lose the opportunity to sell their shares at a premium above the current market price."

Anthropic plans to use five different classes of stock. Among them, Class T shares will give a long-term interests trust the right to select directors in stages, enabling it to select a majority of the board within four years at the latest; Class F shares will enable the seven founders, through entities they establish, to collectively control 50.1% of shareholder voting power.

The public release of this prospectus means that Anthropic is expected to complete its listing after the U.S. midterm elections on November 3 and before the Thanksgiving holiday on November 26, which may ease investors' anxiety.

As Anthropic accelerates its IPO, the AI industry is also at a critical juncture. On the one hand, fierce debate over the safety of this technology continues; on the other hand, as the AI investment boom becomes increasingly reliant on debt financing, the market is increasingly worried about whether these investments will ultimately pay off.

Just over a year ago, Anthropic was valued at $183 billion; in the latest private funding round in May this year, the company's valuation rose to $965 billion.

As market conditions become more severe, investors' expectations for Anthropic's listing valuation are also changing. According to the UK's Financial Times, existing investors expect the company may seek a valuation of $1.5 trillion to $1.75 trillion. In August this year, several Anthropic investors said they expected the company's listing valuation to reach $2 trillion or even higher.

Anthropic believes that AI is expected to "improve productivity across all areas of the economy." But it also warned that its models' "ability to evade systems designed to ensure they act as intended" is increasing, and the company has recorded cases of its models "manipulating or evading" monitoring mechanisms.

In terms of competition, the only competitors Anthropic mentions in the prospectus are OpenAI and Google, as well as "well-funded large technology companies that have reached the frontier through investment." In addition, Anthropic also faces competition from "a broader range of model providers," including open-weight or open-source models.

Responsible editor: Zheng Jingxin Image editor: Chen Feiyan Proofreader: Liu Wei The Paper news tip line: 021-962866 The Paper, unauthorized reproduction prohibited +1 Favorite I want to report #Anthropic #AI #ArtificialIntelligence View more

Source

The Paper