Anthropic IPO Prospectus Reveals Deep Dependence on Big Tech Partners
Anthropic's confidential IPO prospectus has provided investors with one of the clearest looks yet at the business economics behind one of the world's most valuable artificial intelligence companies — and it shows just how closely the AI startup is tied to the Big Tech companies helping it scale.
The Claude maker generated about $4.6 billion in revenue in 2025, nearly 12 times its 2024 revenue. But almost half of that 2025 revenue was sold through cloud marketplaces operated by Amazon and Alphabet's Google, according to the prospectus reviewed by Reuters. Those same companies are also major investors in Anthropic and important suppliers of computing capacity.
The filing highlights a central tension in the AI business: Anthropic needs enormous amounts of computing power and access to global enterprise customers, while many of the companies providing that infrastructure are also building competing AI products.
For prospective IPO investors, that creates a business model with extraordinary growth potential but unusually complex dependencies.
Nearly Half of Anthropic's Revenue Runs Through Amazon and Google
Anthropic's dependence on Amazon and Google has increased rapidly alongside its revenue.
Sales through the two companies accounted for 11% of Anthropic's revenue in 2023, rising to 32% in 2024 and nearly half in 2025. About $2.16 billion of its 2025 sales went through Amazon and Google's cloud marketplaces.
Anthropic paid approximately $351 million in distribution fees to the two cloud platforms during 2025. That works out to roughly 16 cents for every dollar of marketplace sales.
The arrangement gives Anthropic access to established enterprise distribution networks. A customer already using Amazon Web Services or Google Cloud can purchase Claude through those platforms without creating an entirely separate procurement relationship with Anthropic.
That can accelerate adoption, particularly among large companies with existing cloud contracts.
But it also means part of Anthropic's economics is controlled by businesses that have their own AI ambitions.
Amazon and Google Are More Than Cloud Providers
The relationship is unusually complicated because Amazon and Google occupy several roles simultaneously.
They are investors in Anthropic, cloud infrastructure providers, distribution channels and, in some areas, competitors.
Amazon has invested billions of dollars in Anthropic and designated AWS as its primary cloud training partner. Google has also invested heavily in the company and provides cloud infrastructure through Google Cloud.
Anthropic's prospectus acknowledges that reliance on a limited number of partners and suppliers can create conflicts of interest. Cloud providers also have visibility into Anthropic's pricing and commercial terms, potentially giving them information that could influence how they allocate computing resources or market Anthropic's products.
This creates an unusual dynamic.
Anthropic needs Big Tech to expand, but those same companies have the infrastructure, distribution and financial resources to develop competing AI systems.
The Compute Bill Is the Bigger Structural Issue
Revenue distribution is only one side of Anthropic's dependence on Big Tech.
The company also requires enormous amounts of computing capacity to train and operate Claude.
Anthropic reported $54.6 billion of non-cancellable hosting and computing commitments at the end of 2025. By early 2026, its total long-term commitments had risen above $417 billion, covering approximately 3.5 gigawatts of dedicated computing capacity.
Reuters separately reported that Anthropic expects to spend at least $518 billion over a decade on AI infrastructure with six partners. The figure ranks among the largest infrastructure commitments associated with an AI company.
These numbers show why the AI business cannot be evaluated purely on revenue growth.
An AI model can attract customers rapidly, but serving those customers requires GPUs, specialized chips, data centers, networking, electricity and cloud capacity.
The economics therefore depend on whether revenue grows quickly enough to cover the enormous infrastructure obligations required to generate it.
Anthropic's Revenue Growth Is Impressive — But Losses Remain Large
Anthropic's top-line expansion has been extraordinary.
Revenue increased almost twelvefold in 2025 to roughly $4.6 billion. About $3.8 billion came from customers paying based on their usage of Claude, while subscription revenue was about $789 million. The company expects consumption-based revenue to remain the substantial majority of its business for the foreseeable future.
But rapid revenue growth has not eliminated the company's losses.
Anthropic's operating loss exceeded $8 billion in 2025, according to Reuters' review of the prospectus. Its net loss was substantially larger because of accounting effects associated with the valuation of financing instruments.
That distinction matters for IPO investors.
A rapidly growing revenue line does not necessarily mean the company is generating enough cash to fund its infrastructure requirements. Anthropic's future economics will depend heavily on improving margins, increasing customer usage and controlling the cost of compute.
Cloud Providers Also Handle a Growing Share of Cash Collection
Anthropic's reliance on its cloud partners extends into collections.
At the end of 2025, the company had $909 million in outstanding customer payments. Cloud providers were responsible for collecting about 60% of that amount, up from 42% a year earlier.
Anthropic warned that disputes or delays in this payment pipeline could hurt cash flow even when the underlying customer relationship is directly with Anthropic.
For a company spending billions on infrastructure, cash conversion is important.
If revenue is growing rapidly but cash collection becomes slower or more dependent on intermediaries, the company can face additional working-capital pressure.
Customer Concentration Adds Another Risk
The prospectus also highlights concentration on the customer side.
Two unnamed customers each accounted for 12% of Anthropic's revenue in 2025. The company warned that many of its largest customers are not locked into long-term contracts and could reduce or stop their spending.
This is particularly relevant because much of Anthropic's revenue is consumption-based.
Under a usage-based model, customers pay according to how much they use the AI service. That can produce rapid revenue growth when usage rises, but it can also expose the company to sudden changes in customer behavior.
If businesses reduce AI experimentation, shift workloads to cheaper models or switch providers, Anthropic's revenue could be affected relatively quickly.
Anthropic's Accounting Method Will Face Investor Scrutiny
Another important issue is how Anthropic reports revenue from cloud marketplaces.
Anthropic records the full value of marketplace contracts as revenue because it considers itself the principal in those transactions. The cloud providers' share is then recorded as an expense.
This produces a higher reported top line than a method that records only Anthropic's net share.
Reuters reported that OpenAI has argued this approach makes Anthropic's revenue appear larger than it would under a different presentation. Anthropic has defended its accounting, saying it follows established accounting practices and that it is the principal because it sets prices and delivers the service.
Investors will therefore need to look beyond headline revenue when comparing Anthropic with other AI companies.
The more useful questions include how much revenue is generated after marketplace fees, what gross margins look like after computing costs and how much cash the company retains from each dollar of AI usage.
Big Tech Dependence Is Also an Advantage
The risks should not obscure why Anthropic has chosen this model.
Building a global cloud infrastructure network independently would require enormous capital and years of development.
By distributing Claude through Amazon, Google and Microsoft, Anthropic can tap into existing enterprise sales channels and cloud infrastructure. That allows the company to reach customers that would otherwise be difficult for a relatively young AI company to acquire directly.
Anthropic itself describes these partnerships as strategically important for expanding market penetration.
The same relationships can therefore be viewed as both a competitive advantage and a dependency.
For investors, the question is whether Anthropic can eventually build enough direct customer relationships and infrastructure flexibility to reduce its exposure to any single partner.
What the Anthropic IPO Means for Amazon, Alphabet and Other AI Stocks
The IPO could also provide useful information for investors in major technology companies.
For Amazon, Anthropic represents both an investment and a source of AWS demand. More Claude usage can translate into greater cloud consumption, although Amazon also bears the cost of providing the infrastructure and participates in Anthropic's distribution economics.
For Alphabet, Anthropic provides additional demand for Google Cloud and its AI infrastructure while simultaneously competing with Google's own Gemini products.
The arrangement also illustrates the broader AI infrastructure cycle. Hyperscalers can benefit from selling computing capacity to AI startups even when those startups compete with their own AI models.
That creates an important distinction between AI model competition and AI infrastructure economics.
A company does not necessarily need to win the AI-model race to benefit from rising demand for compute.
What Investors Should Watch Before the IPO
Anthropic's IPO prospectus gives investors several areas to monitor:
Revenue growth: Whether rapid expansion continues as the customer base becomes larger.
Gross margins: Whether improving AI efficiency can offset rising compute costs.
Cloud marketplace dependence: Whether the share of sales routed through Amazon and Google keeps increasing.
Infrastructure commitments: How Anthropic manages its hundreds of billions of dollars in long-term obligations.
Customer concentration: Whether major customers sign longer-term contracts or remain usage-sensitive.
Cash generation: Whether revenue growth begins translating into sustainable operating cash flow.
Direct sales: Whether Anthropic can build more direct relationships with enterprise customers.
Competitive pressure: How Claude performs against OpenAI, Google, Meta and other AI systems.
IPO valuation: Whether the eventual public-market valuation properly reflects both growth potential and infrastructure obligations.
Anthropic confidentially submitted a draft Form S-1 to the U.S. Securities and Exchange Commission in June 2026, but the number of shares and offering price had not been determined at that stage.
Conclusion
Anthropic's IPO prospectus reveals a company growing at extraordinary speed but operating within an ecosystem dominated by the same Big Tech companies it relies upon for cloud infrastructure, distribution and financing.
Nearly half of its 2025 revenue flowed through Amazon and Google, while its long-term computing commitments had climbed above $417 billion. At the same time, Anthropic was still reporting operating losses of more than $8 billion.
For prospective investors, the central issue is not simply whether Claude can continue gaining market share. It is whether Anthropic can convert that growth into durable margins and cash flow while managing enormous infrastructure commitments and reducing the risks created by dependence on powerful partners that are also competitors.
The eventual IPO filing, offering valuation and updated financial statements will provide the next major test of whether the public market is willing to price Anthropic primarily as a high-growth AI company or as an infrastructure-intensive business with substantial execution obligations.
Follow AIViralHub for more updates on Anthropic, AI companies, technology stocks, IPOs and global markets.
This article is for informational and educational purposes only and should not be considered investment advice.