Anthropic's epic IPO takes a pivotal step! The IPO application documents will be announced on Monday, and the $2 trillion valuation target is being tested in reality
Zhitong Finance App learned that Anthropic is preparing to take a critical step towards opening up capital markets at a time when the artificial intelligence (AI) industry is still being subjected to multiple interrogations of capital expenditure, commercial returns, and technological security. According to reports, Anthropic will release its initial public offering (IPO) S-1 documents at 9 a.m. EST on Monday (21:00 Beijing time). The company will officially launch an IPO promotion campaign as early as the week of November 9. According to people familiar with the matter, the company expects the size of its IPO to be comparable to that of SpaceX (SPCX.US), and possibly surpass the latter.
What do you read in the prospectus? Revenue quality, cash consumption, and computing power commitment are the three keys
Over the past few weeks, draft IPO filings and related reports have successively leaked some Anthropic financial information. Therefore, some of the disclosures in the S‑1 file may not be too surprising. But the upcoming S-1 document may clarify a lot of vague situations.
For investors, Anthropic's IPO application documents should not only look at revenue growth rates and valuation figures, but more importantly, dismantle the sources of growth, the cost structure, and the costs required to maintain growth. In particular, it is important to note that a company's rapid revenue growth does not automatically mean that its cash flow has improved, let alone that a high valuation is necessarily reasonable.
First, investors need to pay attention to the speed and composition of revenue growth. Anthropic's commercialization mainly revolves around the Claude model. The business includes not only the use of the model by enterprises and developers, but also revenue streams such as subscriptions. The pay-as-you-go model allows revenue to expand as customer calls grow, but its sustainability also depends on how often customers use it, how much to reason, how much to price, and whether customers will switch to other vendors as the model becomes more competitive.
Second, it is necessary to distinguish between book revenue, gross margin, and actual cash recovery. AI model companies must continue to invest in expensive computing power resources, R&D talents, and product development. While revenue is growing rapidly, costs may also rise at the same time. Investors should focus on observing whether the ratio of inference costs to revenue declines, whether gross margin can continue to improve, and when operating cash flow and free cash flow are likely to be corrected.
Third, the procurement of computing power and long-term contract commitments in the prospectus are particularly important. AI companies need to lock down data centers, cloud computing resources, chips, and electricity supplies in advance to avoid capacity bottlenecks as demand grows. But long-term commitments also mean that even if customer demand falls short of expectations, the company may still incur huge expenses. Therefore, investors need to determine whether these contracts are strategic assets to guarantee growth, or whether they may turn into a heavy financial burden when the industry cools down.
Finally, governance structures, founder control, customer concentration, supplier dependency, and AI security risks cannot be ignored. Anthropic's technological strengths and corporate customer relationships may constitute competitive barriers, but if key customers, cloud service channels, and computing power supply are concentrated in a few large technology companies, their business growth may be significantly affected by partner business decisions.
The US stock IPO was “frozen” in October, and all Wall Street is waiting for Anthropic to break down
In the first half of 2026, the US stock IPO market experienced a wild run. According to data compiled by Bloomberg, US stock IPOs had raised up to $162.6 billion by the end of September. However, the style of painting changed abruptly in the fourth quarter. Since the end of September, the US IPO market (excluding shell companies and other financial instruments) has raised only $587.7 million. This directly led to October becoming the slowest month to raise capital for US IPOs throughout the year.
The direct reason for the stagnation of the US stock IPO market points to AI infrastructure companies. Companies lining up to go public in the US, such as Nscale, DayOne, and Data Centers, are still struggling to sell, but investors' appetite for risk has evaporated — as peer shares like CoreWeave are plummeting.
The deeper reason is the resurgence of the “valuation gap” (Valuation Gap).
After a promising first half, there was a huge gap between the high valuations expected by corporate management and the price investors were willing to pay. Will Braeutigam, head of US capital market transactions at Deloitte, described this bid-ask spread as “exploding” and “there is an impasse in valuation, and this gap must be bridged before the wider market moves forward.”
In the face of a stagnant market, Wall Street is betting all of its hopes for a revival on Anthropic's anticipated listing. People familiar with the matter said that Anthropic is expected to launch next month, and its financing scale may reach or exceed SpaceX's listing scale. This is not only about whether Wall Street can have a record “good year,” but also to “clear the wall” for companies that have a wait-and-see attitude in the future.
However, with the exception of Anthropic, the market no longer has high expectations for the rest of the targets in the short term. West Riggs, head of equity capital markets at Truist Financial Corp., said, “We have a lot of optimism about 2027, especially in the software sector, but from an IPO perspective, the overall feeling is quite sluggish. Even a few non-AI projects we're working on seem to have been delayed.” Will Braeutigam also predicts that there will be no more than 6 companies that have raised more than 500 million US dollars in the US stock market before the end of the year. “Most of our customers are focusing on the second quarter of next year.”
The $2 trillion valuation faces re-examination
Anthropic's most popular valuation target is undergoing a process from optimistic expectations to realistic tests. Previously, the market's estimate for Anthropic's IPO reached about $2 trillion. As a comparison, the company was valued at around $965 billion when it completed the financing round in May 2026. If it goes public at a valuation of $2 trillion, it means that its valuation will double within a few months. This valuation also means that investors need to believe that there is still huge room for growth in revenue, market share, and long-term profitability.
But as more financial data surfaced, investors' attitudes began to diverge. According to reports, after Anthropic's latest round of investor communication, some existing investors and potential IPO investors began discussing a valuation of about 1.5 trillion US dollars, which is lower than the previous market forecast of about 2 trillion US dollars. One reason investors are concerned is that the company's revenue is still growing, but the quarter-on-quarter growth rate has slowed.
According to the data, Anthropic's third-quarter revenue reached US$16.8 billion, achieving adjusted profit for the second consecutive quarter. However, the month-on-month revenue growth rate fell back to 45% from over 140% in the second quarter.
In the tech industry as a whole, the 45% month-on-month increase in quarterly revenue is already quite an impressive achievement, but the problem is that Anthropic has grown so fast before. Anthropic's revenue in the first quarter of this year was 4.7 billion US dollars. In the second quarter, it rapidly grew to 11.6 billion US dollars, an increase of nearly 147%. By the third quarter, although revenue had increased by another 5.2 billion US dollars, the growth rate had clearly slowed. Wall Street's previous estimate of Anthropic's valuation of up to 2 trillion US dollars was based on this ultra-rapid growth.
As a result, after quarterly revenue growth slowed month-on-month, some investors began to recalculate their accounts. According to three investors attending Anthropic Investor Day, some investors have already begun discussing a listing valuation of around $1.5 trillion. Jared Carmel, managing partner of Manhattan Venture Partners, said in an interview that Anthropic's current IPO must set a price that investors are willing to accept. He anticipates that Anthropic may eventually go public at a valuation of $1.4 trillion to $1.5 trillion.
At the same time, the other side of high growth is huge cost investment. Anthropic's net loss in 2025 was close to 42 billion US dollars, which is more than 5 times the previous year's net loss of about 8.3 billion US dollars. However, about $34 billion of this was due to changes in the fair value of convertible financing instruments — accounting losses due to soaring valuations, not cash expenses. Excluded, the actual operating loss was approximately US$8.06 billion, an increase of 170% over the US$2.98 billion in 2024.
The reason for the increase in losses was the expansion of upfront expenses for computing power. In 2025, spending on computing power and infrastructure reached US$7.33 billion, equivalent to 1.6 times that year's revenue. But the loss rate is narrowing — the operating loss rate fell from around 750% in 2024 to 176% in 2025.
The $2 trillion valuation reflects investors' high expectations for the AI industry's long-term growth space. Compared with traditional software companies, Anthropic not only hopes to earn revenue through model subscriptions, enterprise services, and API interfaces, but also bets that AI can gradually enter core enterprise scenarios such as programming, customer service, knowledge management, and business process automation. As enterprises move from experimental applications to production deployment, there is an opportunity to upgrade AI services from ancillary tools to part of the enterprise infrastructure.
In particular, the expansion of the Claude series model in the field of enterprise applications and programming has enabled Anthropic to seek more sustainable commercial demand. Once enterprise customers embed AI models into software development, internal knowledge retrieval, or daily operation processes, they often need to continue to pay for model calls and services. This kind of demand is expected to generate high revenue stickiness, and also provide support for the company's future revenue growth.
However, high valuations also mean higher fulfillment requirements. Investors are not only buying Anthropic's current size of business, but also paying for its growth expectations over the next few years. If revenue growth slows down after listing, profit improvements fall short of expectations, or if the AI industry's overall valuation shrinks, the company's stock price may face significant pressure. Therefore, whether a $2 trillion valuation can be established ultimately depends on whether the market believes in Anthropic's ability to transform current technological leadership and demand growth into long-term cash flow generation capabilities.
$518 billion infrastructure commitment! It is an engine of growth and a source of financial stress
If revenue growth is at the core of Anthropic's valuation story, then the huge demand for computing power and infrastructure is one of the cost factors that need to be examined most in this business model.
According to reports, according to Anthropic's IPO prospectus seen by the media, the company plans to assume approximately $518 billion in cloud computing and infrastructure-related expenses or commitments over the next few years. This scale shows that the company's judgment on future AI needs is very aggressive, which also means that its commercial expansion will be deeply tied to cloud computing services, data center construction, and advanced chip supply.
It is important to note that this figure is a commitment to related expenses and infrastructure for many years to come. It does not mean that the company needs to pay the full amount in one time in a certain fiscal year, nor can it simply be equated with capital expenses or cash outflows that have already occurred. But even when considering differences in performance times and contract structures, such long-term commitments can still have an important impact on a company's future financial flexibility.
There is a clear difference between the cost structure of AI companies and traditional software companies. After completing product development for traditional software, the marginal delivery cost for adding a new user is usually relatively limited; however, large model services require continuous inference calculations. The more frequent user calls and the more complicated the model processing tasks, the higher the related computing power costs. For products that can perform complex tasks over a long period of time, such as programming agents, increased user usage may not only bring more revenue, but also drive up the consumption of computing resources.
Therefore, whether Anthropic can establish attractive profit margins in the future depends on whether multiple factors can be improved at the same time: whether model reasoning efficiency can be improved, whether unit calculation costs can be reduced, whether corporate customers are willing to accept higher prices, and whether the company can reduce computing power procurement costs through long-term contracts and resource scheduling.
If the commercial value of AI models continues to increase, and enterprise customers are willing to pay for higher productivity, then huge infrastructure investment may become a necessary condition for establishing competitive barriers. Conversely, if model capabilities gradually converge, service prices continue to decline, or if enterprise customers begin to reduce AI budgets, long-term computing power commitments may become a heavy cost burden.
This also means that the market cannot simply view the $518 billion infrastructure commitment as a sign of benefit to the chip and data center industry. For upstream suppliers, it may represent future demand; for Anthropic itself, it also represents performance obligations, resource dependency, and future profit pressure.
With listing valuations already heavily dependent on future growth, Anthropic needs to prove to investors that these expenses are not simply investments chasing model size, but long-term investments that can bring sufficient commercial returns.
Customer concentration and cloud platform dependency or limit revenue quality
Another important risk for Anthropic comes from the revenue structure. According to Anthropic's prospectus seen by the media, the company received 2.16 billion US dollars in revenue through Amazon and Google's cloud platforms in 2025, accounting for 47% of the annual revenue. This ratio was only 11% in 2023 and rose to 32% in 2024, indicating that the company's dependence on cloud platform channels is gradually deepening.
For Anthropic, the cloud-based sales model can help the company significantly reduce the difficulty of acquiring customers. Many large customers purchase Claude services directly through the cloud marketplace, which can reduce contracts, settlement, and technology deployment. Anthropic stated in its prospectus that these cloud platforms have sales networks that are difficult for a single enterprise to independently replicate, which helps Claude enter the global enterprise market faster.
The problem, though, is that Amazon and Google aren't ordinary channel providers: they both invest in Anthropic, sell the computing resources needed for training and inference, and have their own AI models and cloud computing businesses. This means that these competitors can access some of Anthropic's pricing and commercial terms, and may affect computing power allocation, product promotion, and customer relationships. Anthropic also acknowledged in its prospectus that reliance on a few partners and suppliers could cause conflicts of interest and adversely affect the company's access to computing resources.
A high customer concentration doesn't necessarily mean there's a problem with the business model. For AI companies in the rapid expansion stage, large technology companies, enterprise customers, and cloud platform partners can often provide a considerable procurement scale in the early stages, which also helps push products into a wider range of application scenarios.
However, when a small number of customers contribute a large percentage of revenue, the company's sensitivity to customer procurement decisions also increases. Anthropic's revenue growth and profit margins may be affected once key customers adjust procurement budgets, switch to self-developed models, reduce call volumes, or demand lower prices in contract negotiations.
What's more complicated is that some of the big partners in the AI industry are both customers and computing power suppliers or investors. While this relationship helps Anthropic gain access to computational resources and commercial channels, it may also cause the company to face some degree of dependency. If partners change commercial terms, adjust resource allocations, or use more resources to promote their own models, Anthropic will need to deal with potential cost and competitive pressures.
The cloud platform channel also involves the difference between revenue recognition and financial benefits. Sales revenue obtained through partners can expand an enterprise's customer reach, but a portion of it needs to be paid to cloud service providers. As a result, investors need to pay attention to revenue, gross profit margin, customer renewal rate, and customer retention after deducting channel sharing, not just the unadjusted operating rate of revenue.
After listing, Anthropic's ability to continue to expand its direct corporate customer base, reduce its dependence on a small number of customers, and improve the diversity of revenue sources will be an important indicator for measuring the resilience of its business model.
Overall, Anthropic's upcoming IPO is at a critical stage where the AI industry narrative shifts from “capability breakthroughs” to “commercial returns.” On the one hand, Claude and its related products drive rapid revenue growth, and the demand for AI tools from enterprise customers provides the company with a realistic business foundation. If the listing goes well, Anthropic is expected to gain access to new funding channels and establish more transparent valuation references for the entire industry.
On the other hand, huge operating investment, long-term computing power contracts, customer and supplier concentration, and AI security risks all mean that its valuation cannot be supported by forward revenue forecasts alone. What the market ultimately needs to see is whether revenue growth can continue to be converted into profit and cash flow, and whether the company can control costs and risks while continuously enhancing model capabilities.
Therefore, what is most interesting about Anthropic's IPO is not whether it can set a new listing scale record, but rather how the open market will price the growth potential, capital consumption, and security risks of cutting-edge AI at the same time. If it can prove that high investment is being transformed into sustainable commercial returns, this IPO may become an important milestone in the maturity of the AI industry; if it fails to meet long-term growth expectations, it may also become a turning point for the market to re-examine AI valuation, capital expenditure, and financing models.