<style>.blog-template-wrapper{width:100%;max-width:1040px!important;padding:0!important;margin:auto}.blog-template-wrapper .w-richtext{width:100%!important;max-width:none!important;padding:0!important;margin:0!important}.blog-template-wrapper h2{font-size:34px!important;line-height:1.25!important;margin-top:48px}.blog-template-wrapper p{font-size:19px!important;line-height:1.7!important;margin-bottom:24px!important}.div-block-223 h2{font-size:34px!important;line-height:1.3!important}@media(max-width:767px){.section.articles .base-container{padding-left:24px!important;padding-right:24px!important}.blog-template-wrapper h2,.div-block-223 h2{font-size:27px!important}.blog-template-wrapper p{font-size:17px!important}}</style><div class="w-richtext tex2jax_ignore mathjax_ignore"><p>Anthropic, the company that created Claude, is about to take part in one of the biggest events in financial markets: going public. Founded in 2021, it has grown from a small AI startup to one of the most valuable private technology companies, and its upcoming IPO could value it at more than $2 trillion.</p><p>For investors, however, what is of real interest is not merely the level of the valuation but whether or not Anthropic will be able to turn the strong demand for artificial intelligence into a profitable business.</p><p>The figures show how much more complex this issue is than it first appears. Anthropic earned almost $4.6 billion in revenue in 2025, about twelve times the previous year, and its annualized revenue continued to grow quickly in 2026. However, the company spent about $7.3 billion on computing and infrastructure in 2025 and posted an operating loss of more than $8 billion. It is precisely because of its combination of very rapid growth and very high costs that Anthropic stands out for its market potential.</p><h2>A Company Growing at an Unusual Speed</h2><p>Anthropic has seen remarkable revenue growth; in one year, its revenue rose from about $380 million to $4.6 billion. At the same time, the company has rapidly expanded its customer base among large businesses that use Claude for coding, research, customer service, and other professional activities.</p><p>This focus on businesses is important because an enterprise customer can be worth much more than an individual user. While a consumer may pay about $20 a month for an AI subscription, a large company can spend millions each year integrating AI into its operations. Anthropic thus wants Claude to be more than just another chatbot; its long-term aim is to have AI become part of the way companies actually operate, whether it's writing software, analyzing information, or automating routine tasks.</p><p>This raises an important question for investors regarding how much companies will eventually be willing to pay for AI.</p><p>If businesses end up relying heavily on Claude, then Anthropic could acquire considerable pricing power. On the other hand, if the models offered by Anthropic, OpenAI, Google, and the other competitors become more and more similar, customers might find it easy to switch from one to another, which would make it very difficult for any single company to charge a premium. The answer might have a significant effect on Anthropic's future margins.</p><h2>The Cost Behind Every Answer</h2><p>A major difference between AI companies and conventional software businesses lies in the cost of actually producing the product.</p><p>In a conventional software company, product development can be costly, but after creation, selling to another customer adds only a small cost. With AI, it is different because each interaction requires computing power.</p><p>To generate a response to a question, Anthropic must use servers and specialized chips. Training a new model also requires more computing power, and more advanced models usually need more resources for both development and operation.</p><p>This means Anthropic's rapid growth involves a physical cost.</p><p>The company must purchase chips, build data centers, supply electricity, and maintain the infrastructure. Although AI products appear on our screens as software, they rely on a massive physical infrastructure behind them.</p><p>That is the reason why Anthropic's spending of $7.3 billion on infrastructure is so significant. The company is not merely spending the money in order to expand its sales team or promote Claude. A major portion of its spending is attributable to the fundamental cost of getting the product to work.</p><h2>From Revenue to Profit</h2><p>We now come to one of the most important aspects to consider in relation to Anthropic's IPO: margins.</p><p>Revenue shows how much money a company gets from its customers, while profit is what's left after it pays the costs needed to generate that revenue. What Anthropic needs from its investors is for them to believe that its revenue will one day increase at a faster rate than its costs.</p><p>It might appear to be an extraordinary achievement if Anthropic's revenue rose from $5 billion to $50 billion. Yet if its costs also rise from $7 billion to $55 billion at the same time, the company would still lose money despite growing revenue tenfold.</p><p>A more attractive possibility is that Anthropic will become more efficient as it grows; if the company is able to raise the amount that customers are willing to pay while at the same time reducing the cost of producing each AI response, its margins could rise considerably.</p><p>That is the reason why the cost of computing is so important.</p><p>If the costs involved in producing AI keep on falling while demand keeps on rising, Anthropic may one day go from being a company which uses up huge amounts of capital to one that produces substantial cash flow.</p><h2>Who is financing the use of the computers?</h2><p>Anthropic is not operating by itself; it has strong partnerships with some of the largest technology companies in the world, such as Amazon and Google, and Nvidia provides a large portion of the hardware used in the AI industry.</p><p>An unusual relationship is thus established between competitors and suppliers.</p><p>Both Amazon and Google are engaged in the AI market themselves and at the same time supply Anthropic with infrastructure; Anthropic thus has to rely on companies that are its partners, investors, and possible competitors as well.</p><p>This dependence becomes even more important when looking at the amount of infrastructure required to support future growth. Anthropic's IPO documents show that the company has made major commitments to future infrastructure, stating it has about $20.3 billion in cash and around $518 billion in future infrastructure obligations.</p><p>These figures clearly show how capital-intensive AI competition has become.</p><h2>What Is $2 Trillion Actually Buying</h2><p>At a valuation above $2 trillion and roughly $65 billion in annualized revenue, Anthropic would trade at about 30 times revenue. That’s a very high multiple: investors are paying not for current performance, but for what the company could become over the next several years. This is common among fast-growing technology companies, but it also increases the risk of disappointment.</p><p>To justify the price, Anthropic must keep growing quickly, maintain its position against competitors, and eventually become much more profitable. If growth slows while costs remain high, the valuation could quickly become difficult to defend.</p><p>This is why revenue growth alone can be misleading: a company can grow very quickly without creating value for shareholders if each additional dollar of revenue costs nearly another dollar to generate. The bull case is that cheaper AI will not reduce spending but expand it. Companies could scale from automating a few thousand customer interactions to millions, or from a handful of employees using Claude to the entire organization. Each request becomes cheaper, but the number of requests could multiply.</p><p>The catch is that heavier usage demands more computing power, which explains why AI companies are investing so heavily in infrastructure.</p><h2>What Could Go Wrong?</h2><p>The bullish case for Anthropic is relatively straightforward. AI adoption continues to increase, Claude becomes deeply integrated into businesses, computing becomes more efficient, and Anthropic's revenue eventually grows much faster than its costs.</p><p>But there are also several risks.</p><p>One is competition. If Claude, ChatGPT, Gemini, and other models become increasingly similar, customers may have less reason to remain loyal to one provider.</p><p>Infrastructure costs are also a factor. To stay competitive and develop the next generation of models, Anthropic may have to keep spending huge sums of money.</p><p>Investors may also have already priced in too much future growth.</p><p>At a $2 trillion valuation, the market is effectively assuming that Anthropic will become one of the most economically important technology companies in the world.</p><p>That is a very high expectation for a company that was founded only five years ago.</p><h2>Why the IPO Matters</h2><p>That is why Anthropic's IPO involves more than just one company.</p><p>It could become a test of how public markets value the entire AI boom.</p><p>Over the last few years, investors have committed enormous amounts of capital to AI companies, chips, and data centers, believing artificial intelligence will eventually generate far more economic value than it costs to build.</p><p>Anthropic's IPO gives public investors a chance to put a price on that belief.</p><p>If the company receives a valuation above $2 trillion, investors are effectively saying they believe today's enormous spending will lead to much larger future profits.</p><p>If the market values it significantly below expectations, it could suggest that investors are becoming more cautious about the relationship between AI growth and profitability.</p><p>This is also why the IPO will be interesting to watch from a quantitative perspective. Rather than simply asking whether Anthropic is growing, investors can look at how different parts of the business change relative to each other: revenue growth versus computing costs, customer growth versus spending, and eventually cash flow versus capital expenditure.</p><h2>The Number to Watch</h2><p>One of the most important things to watch when Anthropic starts trading will be the relationship between revenue growth and computing costs. If revenue rises 50% while computing costs rise only 20%, the business economics are improving. If both increase by 50%, Anthropic may be growing rapidly without becoming much more efficient. That difference could tell investors more about the company's future than its headline revenue numbers.</p><p>Anthropic has already shown enormous demand for AI. The next challenge is proving that this demand can eventually translate into sustainable profits.</p><p>That's what makes the IPO so interesting.</p><p>For the first time, the market will have to put a public price on one of technology's biggest bets: how much is the future of AI actually worth?</p></div>
