Artificial intelligence is expensive, and Anthropic put a price tag on it. The company said in its IPO prospectus that it plans to spend $518 billion over the next 10 years on cloud and computing power for technology it believes could be more transformative than industrialization, electricity, or the internet.
Anthropic has already locked in long-term cloud computing deals with hyperscalers such as Amazon (AMZN +1.95%) and Alphabet (GOOG +0.22%) (GOOGL +0.35%), but investors should note a few key takeaways from this major news.
Image source: Getty Images.
The prices of AI components will continue to rise
Anthropic's spending goals indicate a long-term AI tailwind. Not only will Anthropic continue to spend money, but other competitors will also increase their spending. Artificial intelligence is still in its early innings, and none of the hyperscalers want to be left behind.
The entire data center build-out requires memory chips, electricity, physical buildings, specialized cooling systems, and an array of other essentials that don't receive as much attention. All of those inputs will see rising demand, resulting in greater pricing power for their suppliers.
Research firm TrendForce anticipates high bandwidth memory prices will increase by 121% year over year in 2027. Prices have already soared for these foundational chips as demand far outpaces supply.
It's not just chips. AI data center provider Nebius (NBIS +7.44%) has seen its annual contract value per megawatt more than triple in less than a year for short-term capacity deals. A shortage of compute power combined with insatiable demand is enabling neoclouds to substantially raise prices without meeting meaningful customer resistance.
Consumers and businesses are using these products
Claude and other AI models already have real user case studies. Use of these models is becoming as commonplace as a Google search, and revenue is scaling quickly. Anthropic closed out 2025 with $4.6 billion in revenue; as of July 2026, the business's annual revenue run rate had topped $65 billion.
This jump shows how quickly AI products and services can scale. It's this growth rate that makes the $2 trillion market cap it's hoping to achieve with its initial public offering more feasible, but any slowdown in its growth would cast serious doubt on whether it merits that valuation.
Anthropic is still burning through cash, but the hope is that its revenue growth will remain parabolic and eventually produce a profit. As long as new customers continue to sign up and existing customers stick around, the company can realize that objective.
More importantly, Anthropic is an example of broader AI growth and the commercial value of the technology. If it achieves a profitable return on investment, that gives tech giants, enterprises, and other AI labs a green light to increase spending.
Substantial capital is flowing throughout the artificial intelligence industry, and that trend is set to accelerate. It could make many investors wealthy.
