Anthropic's $2 trillion IPO is set at a price lower than the current valuations of AI stocks.

      The figure originates from investors, not the company itself. Six financiers informed the Financial Times that increasing revenue could enable the five-year-old lab to more than double its valuation in an autumn public offering. At a valuation of $2 trillion, it would surpass SpaceX, which went public at $1.77 trillion in June.

      Anthropic isn't fixing parameters on its own. Several investors mentioned that top executives had not agreed on a target, even in private discussions. The projections come from the investors themselves.

      Despite this, the machinery is in motion. According to Quartz, Morgan Stanley, Goldman Sachs, and JPMorgan are spearheading the offering. The involvement of three major banks is not just hearsay; it indicates the transition from investors talking about a listing to an actual listing being developed.

      Two important figures are highlighted. In May, the company reported that its annualized revenue had exceeded $47 billion, which is the latest figure released by Anthropic. Investors now anticipate that this will grow to $100 billion to $120 billion by the end of 2026, marking an increase of more than ten times within the year.

      It’s crucial to understand what annualized revenue means before considering the number. Anthropic prefers this metric, which estimates full-year sales based on recent performance rather than counting a full year’s receipts. This method is valid for a rapidly growing company, but it also means that a strong month can inflate this figure. It’s not the same figure a public company would report.

      Running the numbers on the published figure shows that if Anthropic is valued at $2 trillion with an annualized revenue of $47 billion, it would translate to approximately 43 times its revenue. Anthropic lacks a publicly traded American counterpart for comparison, but companies viewed as AI beneficiaries have been trading at around 55 times revenue this year, as noted by the FT, mentioning Palantir and the cloud group Nebius. Comparatively, a $2 trillion evaluation doesn’t seem excessive; it falls below the market valuations for businesses growing at a significantly slower rate than Anthropic.

      One investor raised the ante, stating, “If Anthropic is growing 800 percent a year, you’d expect them to trade at the very least 30 times revenue,” suggesting this could position the company at a $3 trillion valuation.

      The first significant risk hinges on the American government. The multiple doesn’t account for the ongoing litigation Anthropic faces with the Department of Defense, which has identified the company as a supply-chain risk. In June, the Commerce Department imposed export controls on Anthropic, leading to a temporary halt in the deployment of its two main models, Fable 5 and Mythos 5. Two investors shared with the FT that this temporary ban constrained revenue growth in June, despite a rebound afterward. This regulatory action already affected growth earlier this year, and it also unsettled customers reliant on those models.

      The second risk is that customers are evaluating their expenditures. As per Artificial Analysis, Anthropic’s leading model costs over two and a half times more to utilize compared to OpenAI’s flagship offering, while Chinese open-weight models are significantly cheaper and have markedly improved this year. According to payment data from Ramp, Anthropic gained market share among U.S. businesses last month. However, those same businesses are reportedly “hitting their limit on AI spend,” with many exploring more affordable alternatives. Some organizations have even gone so far as to retract internal directives encouraging maximum AI usage due to rising costs. The uncomfortable reality is that while Anthropic is gaining market share, its buyers are keen on cutting costs, and Chinese models are continually improving.

      As for why the company remains silent, Anthropic has refrained from commenting, and there’s a rationale beyond mere choice. The company filed confidentially with the Securities and Exchange Commission in June, entering a quiet period that restricts what it can disclose about its financial status. Consequently, all circulating figures stem from investors eager for a successful listing. This does not invalidate the numbers but necessitates careful attribution.

      The investments backing those models are substantial. Venture capitalists, sovereign wealth funds, and other institutions have injected just under $100 billion into the company in 2026 alone. Anthropic’s valuation surpassed OpenAI's for the first time in May, reaching $965 billion post-money, with subsequent secondary trades valuing it at $1.2 trillion.

      The argument in favor of the company is compelling. Anthropic has spent the year rolling out models that outperform the competition and marketing them to businesses rather than consumers. One investor, who has also invested in OpenAI and SpaceX, straightforwardly told the FT that “it’s easy to point out challenges.” He continued, asserting that the company remains “first in performance, positioning, and the exposure that investors desire.”

      There is validity in that last statement. A listing of this magnitude hinges partly on investor inclination toward AI exposure, which does not equate to actual fundamentals. This is why the FT frames the