Anthropic faces a massive spending hurdle ahead of its IPO

A milestone in profitability

Anthropic has reached a significant financial milestone as it prepares for its initial public offering (IPO) next week. According to its IPO prospectus, the company recorded an adjusted operating income of $4.08 billion through the first nine months of 2026. This marks the first time the lab has reached profitability on an adjusted operating basis, a streak that continued into the third quarter.

Adjusted operating income is a metric that excludes specific costs like stock-based compensation, noncash charitable contributions, and payments for a class-action copyright lawsuit settlement. By using this metric, the company can show a positive trend that might be obscured by one-time or non-cash expenses. However, the company's GAAP results—which follow standard accounting rules—show a different picture.

Under GAAP rules, Anthropic reported an operating loss of $1.26 billion through the third quarter of this year. The discrepancy between these two numbers is a key point for potential investors to monitor. It suggests that while the core business model is beginning to generate profit, the total cost of doing business remains high.

The hidden cost of compute

A major factor influencing Anthropic's financial health is the cost of computing resources. To secure the power needed to train and run its models, Anthropic has entered into various agreements with partners like Google, Amazon, and Microsoft. These deals allow the company to acquire computing power without paying all the cash upfront, which helps preserve its current capital.

Despite these clever arrangements, the underlying costs of this infrastructure are substantial. As CNBC's Jim Cramer noted, these noncash charges reflect the actual infrastructure costs the company would eventually have to pay in cash. Investors are cautioned not to ignore these figures because they represent the physical backbone of the company's AI capabilities.

The scale of these commitments is massive, totaling approximately $900 billion in future spending. These agreements include a deal with SpaceX worth up to $120.5 billion through 2029. Because these costs are so high, the company must maintain a very high growth rate to ensure that its revenue outpaces its mounting bills.

Projecting long-term margins

Anthropic is attempting to balance these massive infrastructure costs with a sustainable long-term business model. In its IPO roadshow presentation, the company projected that computing and infrastructure expenses would eventually account for a portion of revenue. This projection serves as the foundation for their goal of achieving a long-term adjusted operating margin of roughly 35%.

To hit that 35% margin target, the company needs to maintain the rapid growth it has shown recently. During the first nine months of 2026, Anthropic saw a 1,206% year-over-year revenue growth. While this is a significant jump, the company must continue this trajectory to stay ahead of the $900 billion in commitments it has already signed.

Cramer suggested that the only way for the company to remain viable is to reach a point where it generates hundreds of billions of dollars in annual revenue. If the company cannot scale its income at the same pace as its infrastructure spending, the profitability it achieved in 2026 could become difficult to sustain. This creates a high-stakes environment for the company's upcoming debut on the Nasdaq.

A test for the AI trade

The upcoming IPO of Anthropic is more than just a single company's debut; it serves as a litmus test for the entire artificial intelligence sector. Because so many other companies depend on the infrastructure and models provided by frontier labs, Anthropic's success is tied to the broader market's health. If Anthropic can successfully navigate its spending commitments, it may provide a roadmap for other AI firms.

The company is expected to price its IPO on November 18 and begin trading on November 19. Investors will be looking closely at how the market reacts to the balance between the company's recent profitability and its massive future liabilities. The ability to turn high-cost research into a high-margin business is the central question for the next chapter of the company's history.

Ultimately, the success of this debut will signal whether the current trajectory of AI spending can be converted into a sustainable, profitable industry for the public markets.