A revised filing that changes the scale of the Anthropic IPO story
Anthropic’s revised IPO filing, accepted by the Securities and Exchange Commission on November 2, 2026, adds a crucial third-quarter snapshot to the company’s public market pitch and confirms the working headline of this story: $33.2 billion in revenue over nine months and roughly $895.4 billion in compute-related commitments . The filing is not merely a routine amendment. It supplies the missing scale behind the IPO narrative: Anthropic is now presenting itself as a company with massive enterprise AI demand, but also one whose expansion depends on securing an extraordinary amount of computing capacity.
The company reported revenue of $33.2 billion for the nine months ended September 30, 2026, up from $2.5 billion in the comparable period of 2025 . Associated Press described that as roughly a 13-fold rise, while also noting that profitability remains elusive despite the surge . The third quarter alone produced about $17.0 billion in revenue, compared with about $11.6 billion in the second quarter and about $1.4 billion a year earlier . In plain terms, the filing says Anthropic is no longer a speculative AI lab judged only on model quality or research prestige. It is now asking public investors to evaluate a business already generating revenue at a scale that places it among the most consequential enterprise technology companies.
Yet the same amendment also clarifies the other side of that growth. Anthropic disclosed approximately $895.4 billion in aggregate committed spending under purchase commitments and lease agreements, substantially all related to compute and infrastructure . Axios summarized the disclosure as a surge in contractual spending agreements on AI infrastructure to nearly $900 billion . That figure matters because it reframes the IPO as a test of whether frontier AI companies can convert demand into durable profits while committing far ahead of revenue realization.
Revenue momentum, but a volatile operating profile
The amended prospectus shows Anthropic’s revenue growth still running at a remarkable pace, but not in a straight line toward conventional profitability. For the third quarter ended September 30, 2026, the company reported revenue of about $17.0 billion and an operating loss of $567 million . That reversed the operating income of $974 million posted in the second quarter of 2026 . The filing attributes the swing mainly to $5.0 billion of compute under short-term commitments, which it says currently costs significantly more than comparable compute procured through multi-year arrangements, and to $2.0 billion of non-cash charges connected to its donation matching program .
This is the core tension in the filing. Anthropic’s demand is large enough to force near-term purchases of expensive incremental compute, but those purchases can squeeze operating results precisely when revenue is accelerating. On the company’s own adjusted measure, adjusted operating income rose to $2.5 billion in the third quarter from $2.2 billion in the second quarter, while adjusted operating margin fell from 19% to 15% . That adjusted metric excludes some charges and should not be read as a replacement for GAAP profitability, but it does show management’s preferred view of the underlying trajectory.
Net loss is even more dramatic. Associated Press reported that Anthropic’s net loss for the first nine months was $104 billion, including a $94 billion accounting adjustment related to financing agreements . The Desk noted that the third-quarter net loss was about $23.3 billion, with roughly $21.9 billion tied to changes in the value of convertible notes and similar financing instruments rather than day-to-day operations . For investors, that distinction is important: the losses are real under accounting rules, but the filing also signals that a large portion is linked to financing structures, not simply the cost of running Claude.
The $895.4 billion compute commitment is the defining disclosure
The most consequential new number in the amendment is not quarterly revenue. It is the $895.4 billion of aggregate committed spend for purchase commitments and leases, including short-term commitments and leases that have not yet commenced . Anthropic says substantially all of that amount is related to compute and infrastructure . Axios reported that about $447.5 billion is expected to be incurred before the end of 2030, while substantially all of the remainder is expected between 2031 and 2045 . The filing also states that about $69.8 billion is due in the 12 months following September 30, 2026 .
That number should not be read as money already spent. It is a schedule of commitments and lease obligations that Anthropic expects to incur over time. But it is still economically powerful because it indicates the scale of future capacity the company believes it needs to satisfy demand and continue model development. The Desk put the investor significance succinctly: these totals are obligations that matter both to Anthropic and to the cloud, chip and infrastructure providers on the other side of the contracts .
Anthropic says investing in compute is a prerequisite for long-term growth and that its compute investments are used for research, model development, serving customers and internal usage of Claude . The company also explains that it is shifting from sourcing compute solely through fixed-term cloud agreements toward a hybrid model that includes directly leasing chips and placing them in dedicated facilities operated on its behalf, a structure it calls Anthropic Direct . That strategy is meant to provide long-term economic advantages and more flexibility, but it also changes the company’s cash-flow profile because infrastructure obligations can be incurred well before the associated capacity begins producing revenue .
Enterprise AI demand is becoming an infrastructure race
The filing’s numbers make clear that Anthropic’s IPO is also a referendum on enterprise AI deployment. The company says customers typically use higher-capability Claude models for complex reasoning, coding and agentic workflows, with usage volume and model choice serving as primary drivers of spending . In the amended filing, Anthropic also says customers with $100,000 or more of revenue on a trailing 12-month basis grew from more than 950 as of September 30, 2025, to more than 11,900 as of September 30, 2026 . That customer expansion helps explain why revenue grew so quickly, but it also explains why compute supply has become a central strategic constraint.
This is the key business question behind the IPO. If Anthropic can use its committed capacity to serve expanding enterprise demand at improving efficiency, the $895.4 billion commitment base can be interpreted as a long-term moat: a supply chain built to support model quality, speed and availability. If demand slows, pricing weakens or model efficiency fails to keep pace with infrastructure cost, the same commitments become a fixed-cost burden. The company itself warns that it does not expect revenue to continue growing at comparable rates and cannot assure investors that growth will continue at all . That caution is standard in IPO filings, but it has unusual weight when paired with nearly $900 billion of disclosed infrastructure commitments.
The short-term compute issue is a particularly important signal. Anthropic says its third-quarter operating loss was driven in part by $5.0 billion of compute under short-term commitments, which currently costs significantly more than comparable capacity under multi-year commitments . The company’s argument is that Anthropic Direct and long-term capacity should eventually allow it to backfill expensive short-term compute with more cost-effective infrastructure as dedicated capacity comes online . In other words, Anthropic is telling investors that the near-term margin pressure is partly the price of demand arriving faster than long-term infrastructure.
Liquidity, listing plans and what remains blank
The amended filing also shows that Anthropic enters the IPO process with substantial liquidity. As of September 30, 2026, before IPO proceeds, the company reported $146.6 billion of liquidity, including $95.6 billion of cash and cash equivalents, $36.3 billion of short-term investments and $14.8 billion of availability under its amended and restated credit facility . The company also said it had significant funding support from strategic partners, including up to $77.0 billion in available financing capacity under note purchase agreements and up to $5.0 billion in equity commitments .
Still, the amendment leaves major IPO mechanics unresolved. The Desk reported that the number of shares and the price range remained blank in the filing, meaning the document still did not specify how much Anthropic planned to raise or at what valuation . The company has applied to list its Class A common stock on the Nasdaq Global Select Market under the ticker symbol “ANT” . Until a further amendment supplies the price range and share count, investors cannot fully assess dilution, valuation or the market’s tolerance for Anthropic’s capital plan.
The investment question: growth engine or compute overhang?
Anthropic’s revised IPO filing tells two stories at once. The first is a growth story: revenue of $33.2 billion over nine months, third-quarter revenue near $17.0 billion and rapid expansion in large customers . The second is a resource story: about $895.4 billion of commitments tied overwhelmingly to compute and infrastructure, with nearly $70 billion due in the year after September 30, 2026 . Those stories cannot be separated because frontier AI revenue is increasingly a function of available compute.
For the AI sector, the filing may become a benchmark document. It shows how enterprise AI demand can scale quickly, but also how much future spending may be required to support that demand. Anthropic is effectively arguing that leadership in advanced AI requires not only better models and safer deployment, but also a balance sheet and supply chain capable of absorbing infrastructure commitments that look more like those of a hyperscaler than a traditional software company.
The revised filing therefore strengthens Anthropic’s market presence while sharpening the risk profile of its IPO. The company has demonstrated exceptional revenue growth and a deepening enterprise footprint, but it has also disclosed commitments that will demand sustained execution for years. Public investors will not be asked simply whether Claude is popular. They will be asked whether Anthropic can turn one of the largest compute build-outs in AI into durable, cash-generating enterprise infrastructure.