Anthropic Targets $2 Trillion Valuation in Largest IPO Filing in History

Behind closed doors in preliminary meetings with institutional investors this week, bankers from Morgan Stanley and Goldman Sachs are circulating a single, staggering figure: $65 billion. That is the annualized revenue run rate Anthropic reached at the end of July 2026, a metric the artificial intelligence company is using to anchor what could become the largest initial public offering in history.[1][4]

The San Francisco-based AI lab, maker of the Claude family of large language models, confidentially submitted its draft S-1 registration statement to the U.S. Securities and Exchange Commission on June 1. Now, as the company prepares to flip that prospectus public ahead of an anticipated October listing, the financial architecture of the offering is coming into focus.[1][2]

Anthropic is reportedly targeting a public valuation of $2 trillion. If achieved, the listing would eclipse the $1.77 trillion valuation Space Exploration Technologies Corp. (SpaceX) secured during its own public debut in June 2026. The $2 trillion target represents a massive premium over Anthropic's last private valuation of $965 billion, which was established just months ago during a $65 billion Series H funding round in May.[1]

To justify that multiple, underwriters are pointing to a revenue trajectory that has virtually no precedent in enterprise software. At the end of 2025, Anthropic's annualized revenue run rate stood at approximately $9 billion. By May 2026, it had climbed to $47 billion. When July closed, the figure had surged to $65 billion—a sevenfold increase in just seven months.[1][4]

The raw quarterly figures reflect the same aggressive scaling. Preliminary data for the second quarter of 2026 shows Anthropic generating more than $11.5 billion in revenue. For context, during the same three-month period in 2025, the company brought in just $787 million. This 14-fold year-over-year expansion is largely driven by enterprise adoption of Claude Code and deep integrations with cloud providers.[1][4]

The mechanics of the offering rely on a regulatory pathway established by the JOBS Act, which allows companies to file their Form S-1 confidentially. As defined by the SEC's regulatory framework, the S-1 "contains the basic business and financial information on an issuer with respect to a specific securities offering," which allows the public to "consider the merits of an offering and make educated investment decisions." By filing confidentially, Anthropic's management can negotiate disclosure details and respond to SEC comment letters without exposing sensitive financial data to competitors.[2][3]

For a company scaling as rapidly as Anthropic, the confidential period serves a specific strategic purpose. It allows the firm to lock in multi-year enterprise contracts and finalize compute capacity agreements before its margins and operating costs are subjected to public market scrutiny. The document must only be made public 15 days before the company begins its investor roadshow. Once the S-1 is unsealed, every line item—from the billions spent on Nvidia GPUs to the exact revenue split among its top customers—will be audited and debated.[2][4]

The core tension in the upcoming IPO is the immense capital expenditure required to maintain this growth. Training and operating frontier AI models like Claude requires vast clusters of specialized hardware and gigawatts of electricity. Anthropic has already committed $100 billion to Amazon cloud services over the next decade and pays SpaceX an estimated $1.25 billion monthly for compute capacity through 2029.[1][4]

These infrastructure costs mean that while top-line revenue is exploding, gross margins remain under intense pressure. Public market investors typically demand software-like margins of 70% to 80% for mature technology companies. Anthropic's current margins are estimated to be significantly lower, meaning the company must prove it can reduce the unit cost of inference as it scales, rather than simply burning cash to acquire market share.[4]

To bridge the gap between current costs and the $2 trillion valuation, Anthropic is asking investors to look ahead to 2028. The company is projecting revenue of $190 billion to $200 billion for that year. Pricing the IPO at $2 trillion means investors are paying roughly 10 times that future, unearned revenue—a multiple that assumes the AI adoption curve will not flatten and that open-source alternatives will not commoditize the underlying models.[1][4]

The consequences of this listing will ripple across the entire technology sector. Until now, the generative AI boom has been priced almost exclusively in private markets, where venture capital firms and sovereign wealth funds operate with structural incentives to mark up valuations. An IPO forces a mark-to-market event. If Anthropic holds its $2 trillion valuation, it provides a concrete, audited benchmark for every other AI startup.[4]

Conversely, if public investors reject the pricing and the stock falls sharply in its opening weeks, it could trigger a broad recalibration of AI assets. Private funds would be forced to mark down their holdings, the IPO window for competitors like OpenAI could narrow, and the immense capital flows sustaining the AI infrastructure build-out could face sudden friction.[4]

The next verifiable checkpoint arrives when the SEC completes its review and Anthropic formally unseals the S-1. At that moment, the abstract narrative of exponential AI growth will collide with the rigid reality of generally accepted accounting principles. The market will finally see exactly how much it costs to generate $65 billion in artificial intelligence revenue, and whether that business model is truly worth two trillion dollars.[2][3]