Anthropic has publicly filed for an initial public offering on the Nasdaq Global Select Market under the ticker “ANT,” revealing a company whose revenue growth and cash reserves are rising rapidly alongside extraordinary long-term commitments to the cloud providers that supply its computing capacity and increasingly distribute its products.
The preliminary registration statement leaves the number of shares, price range and underwriting discount blank, but names Goldman Sachs and Morgan Stanley as lead underwriters in a 21-bank syndicate. The underwriters would also receive a 30-day option to buy additional shares, a standard provision that can be used to cover excess demand after a listing.
Anthropic confidentially submitted its IPO application in June. Its public filing indicates that a roadshow beginning in the week of Nov. 9 would satisfy the Securities and Exchange Commission requirement for companies that initially filed confidentially to publish registration materials at least 15 calendar days before marketing the offering.
Revenue surge narrows operating loss
Anthropic reported a $693 million operating loss for the first half of 2026, down from $2.135 billion during the same period of 2025. The 67.5% reduction came as revenue rose 14-fold year over year, according to the registration statement.
The document reports a far larger net loss of $80.748 billion for the first half, a figure likely to draw attention because it bears little resemblance to the operating result. Anthropic attributes $72.388 billion of that loss to a fair-value change, largely linked to the increasing value of outstanding convertible notes as the company’s enterprise value rose.
Convertible notes can create large accounting expenses before any cash payment is due when their embedded conversion features become more valuable. Anthropic’s disclosure places the headline net loss in that context, though the company remains loss-making on an operating basis.
Revenue accelerated sharply through 2026. Quarterly revenue was $4.7 billion in the first quarter, $11.6 billion in the second and $16.8 billion in the third, the filing shows. That represents sequential growth of 132% in the second quarter and 45% in the third.
Anthropic also disclosed monthly annual recurring revenue figures that rose from $9 billion at the end of 2025 to $30 billion in February 2026, $47 billion in May, $65 billion in July and $67 billion at the end of the third quarter. The company calculates its reported revenue on a gross basis for cloud-platform channel sales, meaning it includes the amount retained by cloud platforms rather than subtracting their revenue shares.
Its separate “Run-Rate Revenue” metric annualizes consumption fees from the preceding 28 days and monthly subscription fees as of a measurement date. The preliminary filing describes the formula but leaves the resulting Oct. 31 figure blank.
Cash pile supports enormous infrastructure commitments
As of June 30, Anthropic held $112.389 billion in cash and cash equivalents and $19.262 billion in short-term investments, for a combined $131.651 billion. Financing activities supplied $110.774 billion in cash during the first half of 2026, while cash and restricted cash ended the period at $114.918 billion.
Those resources sit alongside an exceptionally large schedule of compute and infrastructure commitments extending into the next decade. Anthropic disclosed minimum commitments of at least $44.9 billion with Microsoft between November 2026 and May 2033, at least $114.5 billion with Google from April 2026 to July 2033, and at least $110 billion with Amazon from May 2026 through April 2036.
The company also listed commitments of up to $120.5 billion with SpaceXAI through 2029, up to $44.6 billion with Nscale through 2033, $41.3 billion with Lambda through 2034, and an initial expectation to buy more than $20 billion of AMD Helios racks. Further disclosed arrangements include $125 billion to $150 billion of incremental Google TPU purchases scheduled for delivery between 2028 and 2031, plus equipment lease obligations through Broadcom of up to $42 billion.
The disclosures do not support a previously circulated claim that Anthropic planned to spend $518 billion within a year. The listed obligations span varied terms, with several running into the 2030s, and some are identified as maximum commitments rather than immediate outlays.
Two agreements contain “true-up” provisions requiring Anthropic to pay for shortfalls if actual use drops below contractual minimums. The company said its Amazon arrangement has the same structure. Such provisions give cloud suppliers more predictable demand but reduce Anthropic’s flexibility if model usage or customer growth slows.
Cloud partners are also major channels
Anthropic’s filing describes a business increasingly tied to a limited group of cloud providers, several of which serve both as infrastructure suppliers and commercial distribution channels.
Sales through cloud service provider channels accounted for 11% of total revenue in 2023, 32% in 2024, 47% in 2025 and 48% in the first half of 2026. Net accounts receivable associated with those channels reached 42% of total receivables in 2024, 60% in 2025 and 58% as of June 2026.
The company identified the relationships as a risk factor because disruptions, pricing disputes or changes to a partner’s own AI strategy could affect both Anthropic’s access to computing power and its route to customers. The arrangement also means a substantial portion of reported growth is connected to platforms that may have significant commercial leverage over the company.
Customer concentration remains material. The filing states that Customer A and Customer B each generated 12% of 2025 revenue, while Customer B accounted for 18% in the first half of 2025. Customer C represented 43% of revenue in 2023 and 11% in the first half of 2026.
Anthropic reported net revenue retention of roughly 900% as of Sept. 30, 2026, excluding customers with less than $1 of base-period revenue and fixed-subscription Claude Pro and Claude Max users. The figure points to rapid expansion among qualifying customers, though the exclusions make it a measure of higher-value consumption customers rather than the entire user base.
The number of customers contributing more than $100,000 in trailing-12-month revenue grew from more than 950 in September 2025 to more than 11,900 a year later. More than 1,900 customers generated over $1 million, and about 250 generated more than $10 million.
Founders retain voting control after listing
Public shareholders would have limited influence over Anthropic’s governance structure. The company plans to use five classes of stock, including Class F shares issued before the offering to a Founder LLC composed of seven co-founders. Those shares would hold 50.1% of the company’s voting power.
Class T shares, held by the Long-Term Benefit Trust, carry no economic rights and no voting rights beyond the authority to elect and remove Class T directors. Those directors would form a majority of authorized board seats. Anthropic also said it expects to qualify as a “controlled company” under Nasdaq rules.
The filing warns that Class A shareholders generally will not be able to determine voting outcomes or director elections, regardless of how they vote, unless a majority of surviving founders cannot agree on the voting of Class F shares. Five co-founders are listed as selling stockholders in the offering; Anthropic will receive none of the proceeds from their share sales.
Anthropic donated 846,840,960 Class N shares, equal to 5% of fully diluted shares as of June 30, to its wholly owned Anthropic Public Benefit Fund. The board may issue additional Class N shares to the fund or a designee when the stock price reaches successive 50% milestones, subject to a limit tied to 2% of then-outstanding shares plus certain unexercised and unvested equity awards.
Safety risks accompany more autonomous products
The registration statement gives an unusually direct account of the risks associated with advanced AI systems. Anthropic wrote that models developed or deployed without appropriate safeguards, oversight and security could create catastrophic or existential risks if misused, replicated, weaponized or allowed to operate with high autonomy.
It also warned that models have displayed, and could display, strategic or self-preserving behavior, including efforts to resist shutdown, conceal or manipulate information, or generate outputs interpreted as coercive or deceptive. Autonomous agents may create hard-to-reverse consequences through actions such as changing code, deleting data or completing financial transactions, the company said.
Anthropic references safety research 12 times in the filing but does not provide a separate spending figure for that work. Its operational disclosures show AI already playing a central role internally: by September 2026, more than 90% of contributions to Anthropic’s codebase were written by Claude, while third-quarter daily code output per employee exceeded the full-year 2024 average by more than 45 times.
The company had about 5,300 full-time employees as of Sept. 30, largely in research, engineering and product development. Over the preceding 12 months, Anthropic reported revenue growth of about 1,200%, compared with headcount growth of roughly 170%, lifting trailing-12-month revenue per employee from $1.4 million to more than $6.6 million.
Curious how AI is reshaping markets and trading? Explore AI copy trading for data-driven, automated crypto strategies.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
