Anthropic's expected IPO is less a normal listing than a public-market test of AI math. The company has real revenue momentum, but a $2 trillion valuation asks you to pay now for growth that still has to arrive.
Anthropic's IPO story has moved past whether the company can grow. It can. The harder question is what price you put on that growth when the public prospectus is still private and the number being discussed is larger than almost every listed company on Earth.
According to the Financial Times, investors are betting Anthropic could reach a valuation of more than $2 trillion when it goes public, possibly as soon as October. That would put the Claude maker ahead of SpaceX's record-setting debut and turn a four-year-old AI lab into one of the biggest public companies in the market on day one. That's not normal. Even in this AI cycle, it is a very large ask.
The facts underneath the excitement are real enough. Anthropic said on May 28 that it raised $65 billion in Series H funding at a $965 billion post-money valuation, with Altimeter Capital, Dragoneer, Greenoaks and Sequoia Capital leading the round. The same company announcement said its run-rate revenue had crossed $47 billion earlier in May, up from $14 billion when it announced its Series G in February and from about $9 billion at the end of 2025.
That pace is why bankers and investors can talk themselves into almost any number. A business that moves from $9 billion to $47 billion in annualized revenue in roughly five months does not fit neatly into the old software spreadsheet. But run-rate revenue is still a snapshot. It is not audited full-year revenue. It is not cash flow. It is not net income. You should not treat it as if it were all three.
The valuation depends on what you believe about 2026 #
The Financial Times reported that investor models point to annualized revenue of $100 billion to $120 billion by the end of 2026. That is the number carrying the $2 trillion conversation, not the $47 billion Anthropic disclosed in May. If the company gets there, the multiple looks less wild than the headline. If it misses, the gap gets ugly fast.
Here's the thing: Anthropic has earned attention, but it has not yet earned a blank check from public investors. The company confidentially submitted a draft S-1 to the SEC on June 1, and Anthropic's own notice said the number of shares and the price had not been set. That means the market still has not seen the public filing that matters most: revenue recognition, gross margin, customer concentration, compute commitments, losses and cash burn.
Those details matter more for Anthropic than they would for a normal enterprise software company. Claude usage costs money every time customers ask the model to work. The better the product performs, the more compute it can consume. That doesn't make the business bad. It does mean a revenue multiple alone is a blunt tool when you still don't know how much margin is left after the model runs.
Bloomberg reported in June that Anthropic picked Morgan Stanley and Goldman Sachs to lead work on the IPO, with JPMorgan also on the deal. That is exactly the banking lineup you would expect for a record-sized listing. It also tells you something else: Wall Street wants this deal to happen while AI appetite is still hot.
Claude's enterprise base is the real argument #
The best case for Anthropic is not hype around chatbots. It is the way Claude has become a paid tool for companies and developers, especially through products like Claude Code and cloud distribution through Amazon and Google. Fortune noted in June that Anthropic's confidential filing was a signal that the company wanted to beat OpenAI to the public markets, and that its $47 billion run-rate made it hard to dismiss as a rumor-driven startup.
That is the part investors should take seriously. Enterprise revenue usually behaves better than consumer subscription revenue, especially when companies build workflows around a model and start paying for usage at scale. Developers are sticky customers when the tool saves time. Procurement departments move slowly, but once software is inside the budget, it can stay there.
Still, don't confuse stickiness with certainty. OpenAI, Google, Meta, xAI and Chinese model makers are all pushing prices, features and benchmarks. Customers may like Claude today, but they will compare it against whatever model is cheaper, faster or easier to wire into their existing stack next quarter. Frontier AI is not a sleepy software category. It is a knife fight with enormous data center bills attached.
That is why the public S-1 will matter more than another valuation leak. If Anthropic shows strong margins, broad customer demand and a credible path to profit, a giant IPO will look aggressive but understandable. If the filing shows thin margins, concentrated revenue or heavy supplier-linked spending, investors will have to decide how much of the AI boom is real demand and how much is money moving in a circle among model labs, cloud providers and chip companies.
For now, the story is current, concrete and very simple. Anthropic has disclosed $47 billion in run-rate revenue and a $965 billion private valuation. Investors are now talking about more than $2 trillion. The missing document is the one that tells you whether that second number is ambition, discipline or just the market losing patience with ordinary math. Also read: Nvidia Discloses $21 Billion SpaceX Stake And $30 Billion In Intel Shares • Alibaba's Qwen3.8-27B Squeezes Frontier AI Benchmarks Onto One Gaming GPU • China Forces Meta to Unwind Its 2 Billion Dollar Acquisition of Manus AI