On October 26, 2026, Anthropic formally submitted an S-1 registration statement to the U.S. Securities and Exchange Commission, seeking to list its Class A common stock on the Nasdaq Global Select Market under the ticker "ANT." The prospectus shows that the company's revenue in the first half of 2026 reached $16.3 billion, up about 1,300% from $1.16 billion in the same period last year; its net loss on the books for the same period was $80.7 billion, of which $72.4 billion came from non-cash items such as fair value changes in convertible notes, derivative liabilities, and SAFEs. Goldman Sachs and Morgan Stanley co-lead a syndicate of 22 underwriters. The company could begin trading as early as around November 16, with a potential valuation range of $1.5 trillion to $2 trillion.
The Real Meaning of the Revenue Curve: Not Growth, but a Leap
Looking only at the 1,300% year-over-year growth rate can be misleading. What better illustrates the situation is the absolute figures of the company's historical revenue: according to the S-1, Anthropic recorded only about $386 million in revenue for all of 2024 and about $4.6 billion for all of 2025, while revenue in the second quarter of 2026 alone already reached $11.56 billion—one quarter exceeding the total of the past three years. This means that in the 91 days ended June 30, 2026, the company earned more than it did in 2023, 2024, and 2025 combined.
In terms of growth structure, compute and infrastructure costs grew about 300% over the same period, far below the 1,300% revenue growth. This divergence indicates that economies of scale are emerging: the compute investment required for each additional dollar of revenue is declining rather than expanding proportionally. In the second quarter, Anthropic recorded $974 million in operating profit, the first positive figure among the six quarters disclosed in the S-1, achieved right before the prospectus filing.
Anatomy of the $80.7 Billion Loss: Book Numbers and Real Cash Burn
The $80.7 billion net loss in the prospectus will invite considerable misreading and needs to be broken down clearly. According to the S-1, the bulk of this figure is not operating cash expenditure—$72.4 billion comes from accounting treatment of fair value changes in convertible notes, derivative liabilities, and SAFE agreements. Before going public, Anthropic raised a large amount of convertible securities from investors; as the company's valuation continued to climb, accounting rules required these securities to be continually marked up in book value, creating a huge "loss" with no actual cash outflow.
The true scale of operating cash burn is far smaller than $80.7 billion, and in the second quarter the company already turned profitable at the operating level. The absolute magnitude of $80.7 billion is still rare, and it clearly reflects that Anthropic's valuation has made an order-of-magnitude leap over the past two years. In May 2026, the company raised $6.5 billion in its latest funding round at a valuation of $96.5 billion, while the target valuation range for this IPO is as high as $1.5 trillion to $2 trillion.
Another Story Revealed by Enterprise Customer Data
Equally important as the revenue figures is the change in customer mix. According to the S-1, the number of enterprise customers spending more than $100,000 annually grew from about 950 as of September 30, 2025, to more than 11,900 as of September 30, 2026, an increase of about 12.5 times in one year. About 250 enterprises spend more than $10 million annually.
A better indicator of business quality is net revenue retention (NRR): about 900% as of September 30, 2026. This means that Anthropic's enterprise customer base from a year ago is now spending, on average, nine times what it spent last year. For SaaS companies, NRR above 100% is generally considered good, above 150% is top-tier, and 900% has almost no precedent. The mechanism driving this figure is not simply expansion sales; rather, after the Claude API was deeply embedded in enterprise workflows, usage expanded nonlinearly as enterprise AI adoption accelerated.
The Dual Narrative in the Founder's Letter: An Age of Miracles Alongside Existential Risk
There is an unusually information-dense detail in the prospectus: the risk factors section of the S-1 runs more than 90 pages, including the company's explicit acknowledgment that AI could pose "catastrophic or existential risks to humanity." This is not boilerplate lawyerly disclosure, but based on a judgment CEO Dario Amodei has publicly stated—he estimates the probability of AI causing human extinction at between 10% and 25%.
At the same time, the founder's letter in the prospectus describes the company's vision as ushering in an "Age of Miracles" and explains to potential shareholders the enormous value AI will create. In the same document, it must both paint a trillion-dollar business outlook and honestly disclose that the founder believes this technology carries a double-digit probability of existential risk to humanity—a narrative tension with almost no precedent in capital markets. It forces investors, as they read the prospectus, to solve an unprecedented risk-pricing problem: how do you write down the share price you are willing to pay next to a risk factor that says "there is a probability of human extinction"?
The prospectus also discloses the founders' voting control structure: the founding team retains more than 51% of voting power, and outside public shareholders will not be able to dictate the company's governance direction. This "dual-class" structure is not uncommon among technology companies (Google and Meta both use similar designs), but in Anthropic's context it has special significance—it means that even if the share price comes under pressure in the future, management can still act independently according to its own judgment of the safe path, without having to respond to shareholder pressure to accelerate commercialization that could jeopardize safety principles.
Comparison with SpaceX's Record: Can an AI Company Set a New Benchmark?
The market reference point for this IPO is already clear: SpaceX raised $75 billion in its June 2026 IPO, surpassing the $29.4 billion record set by Saudi Aramco in 2019 to become the largest IPO in history. Anthropic's fundraising size may exceed $100 billion; if so, it would break the record SpaceX has just set, only four months after the previous record was broken.
The difference in valuation logic is also worth noting. SpaceX is a company with substantial hardware assets, launch contracts, and government customers, and its valuation has relatively traditional financial anchors. Anthropic's core assets are training data, model weights, and engineering talent; it has no factories, no satellites, and no government launch contracts—its valuation is almost entirely a bet on future compute competition, generational model transitions, and enterprise AI penetration. A $2 trillion target valuation means the market must believe that Anthropic can sustain an NRR close to 900% for a considerable time while also defending its current technology position in competition with OpenAI and Google DeepMind.
Why Go Public Now
From a financial timing perspective, Anthropic's choice to file the S-1 now has its own logic: the second quarter had just delivered the first operating profit, providing an inflection-point story for the roadshow; revenue growth of 1,300% is at a historic high, and waiting longer would enlarge the base, naturally narrowing the year-over-year figure; and among the $6.5 billion financing completed in May 2026, some convertible notes include milestone trigger provisions, and going public may be one of the execution paths for some of those provisions.
The regulatory environment is also a variable. The S-1 specifically lists the policy direction of the Trump administration as a risk factor—its policies could provide a more permissive regulatory environment for AI development, or they could raise the procurement cost of compute hardware through tariff policy. This two-way uncertainty in corporate governance itself reflects the complexity of the current AI policy ecosystem.
Independent Judgment
Anthropic's S-1 is a rare commercial document: it reveals at once both the AI industry's growth gravity and its existential anxiety. The 900% NRR is real, showing that once enterprise customers deeply rely on Claude, they are hard to remove; the second-quarter operating profit is also real, showing that scaling has indeed produced marginal benefits. But the $2 trillion valuation target requires the market to believe these trends will continue with very high probability, while the prospectus itself uses 90 pages of risk factors to remind readers how uncertain that is.
For capital markets, the real challenge posed by Anthropic's listing is not the valuation number but the pricing framework itself: this is a company whose CEO publicly acknowledges a double-digit risk of human extinction, yet it is simultaneously becoming part of global enterprise infrastructure. Regardless of the final listing price, Anthropic's S-1 will become a historical document of the AI capitalization era—a prospectus filled with real data, real contradictions, and real uncertainty. In the history of technology IPOs, that already makes it an honest outlier.
© 2026 Winzheng.com 赢政天下 | 转载请注明来源并附原文链接