Anthropic Wants to Raise Up to $90.4 Billion in an IPO and Beat SpaceX's Record

Anthropic has decided to go public and immediately announced record-breaking scale. According to Bloomberg, Anthropic PBC and some of its shareholders want to raise up to $90.4 billion through an IPO and a private placement. The headline of the Bloomberg piece mentions that large technology companies are taking part in the offering. If all goes well, Anthropic will become the first frontier AI lab on the stock market, and the deal will be larger than SpaceX's record initial share sale.

What is known and what is not yet

There are few facts, and it is important not to speculate. The stated goal is known: up to $90.4 billion. The structure is known: a public share offering plus a private placement. It is also known that securities will be sold not only by the company itself but also by some of its shareholders. This matters: part of the money may go not to training models and data centers, but to those who want to lock in profits.

Still unknown are the company's valuation, the share price, the timing, the makeup of anchor investors, and exactly how much will come from Big Tech. This is a maximum target, not a closed deal. The final amount could turn out to be noticeably smaller.

Why this matters even if you are not an investor

First: frontier models remain one of the most capital-intensive undertakings in the industry. Going public shows that private rounds are no longer enough for the labs. They need the deepest pool of money, and that is the public market.

Second: a public company is required to disclose information. The prospectus will have to show revenue, cost structure, risks, dependence on chip and cloud suppliers, and customer concentration. For developers and product managers building products on top of Claude, this is a rare chance to see the real economics of the API and subscriptions rather than press releases. For example, whether most revenue comes from the enterprise segment rather than from mass-market users. That determines whose needs the lab will serve first.

Third: the pressure regime changes. Quarterly reporting and analyst expectations influence decisions on pricing, limits, and priorities. This is still conjecture, but the market has long seen similar patterns at other companies after their IPOs. It is quite possible that terms for customers will become less "startup-like" and more predictable, or perhaps even tougher.

Fourth: Anthropic is registered as a PBC, a public benefit corporation. It will be interesting to see how this status and its stated focus on safety will coexist with the demands of public investors. The available data does not answer this, but it is one of the main questions for the prospectus.

What to do in practice

If your product depends on a single lab, this is a good reason to reconsider single-supplier risk. You do not have to migrate; it is enough to keep an abstraction layer over models and periodically run your own tests on alternatives. Then a change in price or terms will not become an outage.

Do not build current rates into your business plan as a constant. An IPO of this size means the supplier's economics will be under a magnifying glass, and therefore price lists and limits may be revised more often.

Founders should also consider the market effect. If the deal succeeds, the window for AI companies on the public market will open wider, and the valuations of other players will be compared with Anthropic's result. If it fails, it will signal that appetite for AI valuations has hit a ceiling. Either outcome affects how much money startups around large models will receive.

And one last thing. When the prospectus becomes public, read the sections on risks and customer concentration: they usually contain less marketing and more truth about what your supplier's business rests on.

Limitations

This is a report of intentions, not a completed deal. Without details of the structure and terms, it is too early to draw conclusions about the impact on prices and products. For now it is enough to know that the largest AI lab is going to the public market for a sum comparable to the budgets of entire industries, and that this step will soon force it to disclose figures the industry has not seen before.