Agentic AI,
Artificial Intelligence & Machine Learning,
Next-Generation Technologies & Secure Development
Amazon and Google Marketplaces Handled 47% of the Frontier AI Lab’s 2025 Sales
Recent analyses of the financial standing of the frontier artificial intelligence lab, Anthropic, have highlighted an impressive trajectory of sales growth, contrasted by significant losses. This balance raises questions about the sustainability of its rapid expansion and reliance on major cloud service providers.
According to an initial public offering (IPO) prospectus first reviewed by Reuters, Anthropic, situated in San Francisco, recorded a remarkable 47% of its 2025 sales through the cloud marketplaces of Amazon and Google. This figure signifies a substantial uptick from 32% in 2024 and an even lower 11% in the preceding year. If the IPO progresses as planned, Anthropic stands poised to become the first AI lab in this advanced sector to be publicly traded, setting itself apart from competitors like OpenAI, which is not anticipated to go public until early 2027.
In a striking development earlier this year, Anthropic achieved a staggering valuation of $965 billion, with projections suggesting that its IPO could potentially more than double this figure, speculating a valuation of $2 trillion post-IPO. This meteoric rise, however, is interspersed with risks, as nearly a quarter of Anthropic’s revenue in the previous year stemmed from merely two clients. The absence of long-term contracts with many of its significant customers creates a precarious situation, where spending could be reduced or paused at any moment (as indicated by ongoing discussions among tech giants regarding AI standards).
The financial figures so far for Anthropic have been astounding. In 2025, revenue surged to $4.6 billion—an astronomical 1,200% rise compared to $400 million in 2024, a year the company unveiled its inaugural large language model, along with a notable absence of revenue in 2023. The rapid acceleration continued into 2026, with Anthropic reporting $11.5 billion in sales for just the second quarter, according to insights from The Financial Times.
Delving deeper into the revenue specifics, approximately $3.8 billion was derived from clients utilizing Anthropic’s Claude AI system based on usage. An additional $789 million contributed to the revenue through subscriptions. The prospectus indicates an expectation that consumption-driven revenue will continue to dominate sales in the foreseeable future, with the United States accounting for nearly two-thirds of the company’s overall sales.
Despite these remarkable sales figures, Anthropic isn’t without its financial challenges. The company has reported an alarming increase in net losses, reaching $42 billion in 2025, compared to $8 billion in 2024 and $2 billion the previous year. It’s essential to note that this staggering loss figure encompasses a $34 billion accounting adjustment regarding the increased value of future financing that is projected to become shares. Operating losses also climbed significantly, rising from $2.98 billion in 2024 to over $8 billion in 2025.
Operational expenditures have been steep, with Anthropic investing $7.3 billion in computing and infrastructure in the past year—a figure that rose nearly 300% from $2.5 billion in 2024 and an over 1,800% increase from $400 million in 2023. This expense accounted for more than half of the company’s total operating costs, which stood at $12.65 billion. Furthermore, Anthropic has outlined future commitments of approximately $518 billion for cloud and computing obligations, indicating substantial long-term financial planning.
Both Amazon and Google play a dual role as not just partners for distribution, but also are significant investors in Anthropic. As both companies supply the crucial computing power for the AI operations, they simultaneously serve as direct competitors in the AI landscape. In 2025 alone, Anthropic compensated these tech giants approximately $351 million for $2.16 billion in sales facilitated through their cloud marketplaces, underlining the complex financial interdependencies at play.
By distributing its Claude AI through Amazon, Google, and Microsoft, Anthropic aims to leverage the expansive sales networks of these tech giants, accessing existing customer bases and enhancing market penetration—an endeavor that would be challenging for a single entity to replicate independently. However, such a narrow alliance raises potential conflicts of interest, presenting intricate dynamics that may have implications for the company’s future.
As cash flow becomes increasingly dependent on third-party entities like Amazon and Google, these partners have managed to collect 60% of Anthropic’s outstanding customer payments as of the end of 2025, marking an increase from 42% in 2024. Furthermore, the signing of a cloud computing agreement with Microsoft in late 2025 introduced additional uncertainties regarding cash flow stability, highlighting the risks involved in these arrangements.
In an unusual accounting approach, Anthropic recognizes full revenue from sales agreements involving access to its AI model via marketplace platforms, treating the fees paid to Amazon or Google as marketing expenditures. This controversial method has drawn scrutiny, with competitors like OpenAI alleging that it inflates Anthropic’s revenue figures by billions, prompting investor concerns over financial transparency.