Anthropic is moving toward what could be one of the biggest IPOs ever. Here's what's confirmed

The company that makes Claude is preparing to do something no major frontier AI lab has done: sell shares to the public. If it happens, Anthropic's debut could be one of the biggest IPOs ever, a moment when ordinary investors can finally bet on frontier AI directly. It is also a story where the gap between what is confirmed and what is speculation is unusually wide, and where getting that gap right is the whole point. Here is the careful version.

What has Anthropic actually done?

The confirmed fact is narrow and important: on 1 June 2026, Anthropic announced it had confidentially submitted a draft registration (a draft S-1) to the US Securities and Exchange Commission. That is a first formal step toward an IPO, not the IPO itself.

Crucially, it is a confidential draft, not a public filing. There is no S-1 you can go read on the SEC's website, and the detailed financials inside it have not been released. Anthropic itself was blunt about how little is settled: "the number of shares to be offered and the price have not yet been set," and any offering "will depend on market conditions and other factors." So anyone telling you Anthropic "filed its IPO" is overstating a preliminary, private step.

The $965 billion question

The eye-watering number comes from a private funding round, not the stock market. On 28 May 2026, Anthropic raised $65 billion in a Series H round led by Altimeter, Dragoneer, Greenoaks and Sequoia, at a $965 billion post-money valuation. That figure did something symbolically large: it overtook OpenAI, whose most recent valuation is $852 billion (from a $122 billion round completed in March 2026). For the first time, the maker of Claude is worth more on paper than the maker of ChatGPT.

Handle that number carefully, though. $965 billion is what private investors agreed the whole company was worth in a funding round. It is not an IPO price, a market capitalisation, or the amount Anthropic would raise by listing, all of which are different and, as of now, unset. When the IPO valuation is actually decided, it could be higher or lower.

When will it list?

This is where it shifts from confirmed to reported. According to reporting from the Wall Street Journal, Bloomberg and CNBC, Anthropic has been meeting potential investors and could launch its share sale as soon as this autumn, around September or October 2026, with Goldman Sachs, Morgan Stanley and JPMorgan lined up as underwriters.

But hold the specifics loosely. No IPO date, price, ticker or exchange has officially been set. The widely-repeated "Nasdaq in October" is press speculation, not confirmed, and the actual exchange and ticker will only appear when Anthropic files a public S-1. One thing does make the timing notable: OpenAI, which also filed confidentially, has reportedly pushed its own debut toward 2027, so Anthropic could genuinely be the first big frontier lab to go public, if it lists first.

Is Anthropic actually making money?

This is the most interesting question, because the answer separates Anthropic from its rivals, but only partly. Anthropic's officially disclosed run-rate revenue crossed $47 billion in May 2026, extraordinary growth for a company that was tiny two years ago. Two honest caveats: "run-rate" is an annualised snapshot (a recent month multiplied out), not $47 billion actually booked over the year; and higher figures floating around (roughly $69 to $74 billion by late July) come from unofficial third-party trackers, not the company.

The headline claim is that Anthropic told investors it is on pace for its first profitable quarter, which, if true, would set it apart from OpenAI, still losing tens of billions. But treat this with real caution: it is a claim about a single quarter, not a profitable company (Anthropic's cumulative losses are estimated at roughly $10 to $15 billion), and some analysts argue the "profit" is a timing and accounting artefact rather than a durable turn. So the fair statement is that Anthropic's economics look healthier than OpenAI's, and that this is contested, not settled.

Why it matters

Set the caveats aside for a moment and the significance is real. This would be the first time public investors can own a piece of a frontier AI lab directly, rather than buying it indirectly through Microsoft, Amazon, Google or Nvidia. Given its near-trillion-dollar private valuation, a listing would likely rank among the largest technology IPOs in history. And it would be the first real public-market test of whether the enormous valuations attached to AI can survive contact with ordinary shareholders, quarterly earnings and short-sellers, which is exactly the pressure private rounds never apply.

That is why this listing matters beyond Anthropic. It lands in the middle of the AI-bubble debate: if a near-trillion-dollar AI lab goes public and trades well, it is a powerful vote of confidence; if it stumbles, it validates every bear. Either way, the honest posture right now is to watch the confirmed facts and ignore the invented ones. Anthropic has filed confidentially, is worth $965 billion on paper, and may list this autumn. Everything past that, the date, the price, the exchange, is not yet real. For more, see the AI section and our look at whether the AI boom is a bubble.

Anthropic's IPO, at a glance

Confidential draft S-1	Filed, announced 1 June 2026 (no public S-1 yet)
Private valuation	$965B (Series H, May 2026), overtook OpenAI's $852B
Run-rate revenue	~$47B (official, May 2026; annualised, not booked)
Reported timeline	Possible debut autumn 2026 (Sept/Oct) — not set
IPO price / date / ticker / exchange	None set; "Nasdaq/October" is speculation
Backers / founders	Amazon, Google; Dario & Daniela Amodei (ex-OpenAI, founded 2021)