Anthropic Files for a Nasdaq IPO as ANT: What Its $16.3 Billion in First-Half Revenue — and $80.7 Billion Loss — Actually Mean
Anthropic, the company behind the Claude family of artificial intelligence models, publicly filed for an initial public offering on Monday, 26 October 2026, and applied to list its Class A common stock on the Nasdaq Global Select Market under the ticker symbol ANT.
The filing does not mean Anthropic shares are on sale yet. The registration statement has not become effective, and the company has not set how many shares it will offer or at what price — the two details investors most want to know are still blank.
The filing in numbers: growth first, then the loss explained
The headline figures come from Anthropic’s own Form S-1, as reported by TechCrunch’s filing report and other coverage of the document.
| Measure | What the S-1 shows |
|---|---|
| First-half 2026 revenue | $16.286 billion, up from $1.163 billion a year earlier |
| Second-quarter 2026 revenue alone | $11.56 billion |
| Full-year revenue | $41 million in 2023; $386 million in 2024; $4.59 billion in 2025 |
| First-half 2026 net loss | $80.748 billion |
| Of that loss, non-cash fair-value charge | About $72.4 billion, tied mainly to convertible notes and similar instruments |
| First-half 2026 operating loss | $693 million, down from $2.135 billion a year earlier |
| Computing and infrastructure cost, first half 2026 | $8.945 billion, up from $2.334 billion a year earlier |
That table is why the $80.7 billion figure needs a second look. Most of it is an accounting revaluation, not cash spent running the business in those six months.
When Anthropic’s private valuation rose, the estimated value of convertible notes and related instruments rose with it, and accounting rules recorded that increase as a charge. The operating loss — the closer, though still imperfect, measure of day-to-day business performance in the filing — was $693 million on $16.3 billion of revenue.
Anthropic also warned in the filing that it may not achieve or sustain profitability in the future, and that its revenue growth rate could decline, possibly sharply. Readers should not treat the first half of 2026 as a pace that will automatically continue.
Why revenue jumped so fast
Anthropic’s growth is, on the filing’s own numbers, an enterprise story more than a consumer-subscription story. Usage-based fees, measured in tokens, made up about 80% of first-half revenue, with subscriptions accounting for the rest, according to coverage of the filing.
As of 30 September 2026, Anthropic reported more than 11,900 customers generating at least $100,000 in revenue over the preceding 12 months, up from more than 950 a year earlier. More than 1,900 customers generated at least $1 million, and about 250 generated at least $10 million.
The filing also notes that revenue from a group of existing customers was about nine times higher in the three months ended 30 September 2026 than a year earlier, using a definition that excludes Claude Pro and Max subscribers. That concentration cuts both ways: a small number of large customers can drive extraordinary growth, and losing or renegotiating with them could change the picture quickly.
Computing is the other side of the same coin. At $8.945 billion in the first half, computing and infrastructure cost more than half of revenue. Anthropic trains its models across Amazon Trainium, Google TPUs, Nvidia and AMD chips, and Claude is available on Amazon Web Services, Google Cloud and Microsoft Azure, according to the filing summary.
Who would control Anthropic after it lists?
This is the part of the filing that has drawn almost as much attention as the finances, because public shareholders would not get proportional control.
Anthropic’s seven co-founders — Dario Amodei, Daniela Amodei, Tom Brown, Jack Clark, Jared Kaplan, Sam McCandlish and Christopher Olah — will hold a special Class F share through a founders’ entity carrying 50.1% of general voting power. Separately, Anthropic’s Long-Term Benefit Trust, described in the filing as a financially disinterested trust, will generally have the power to elect and remove a majority of the board.
Anthropic presents that structure as a way to protect its public-benefit and AI-safety mission as it grows. Critics of dual-class and trust structures argue the opposite: that it leaves public investors with little practical ability to change direction if they disagree with management.
The filing also carries extensive risk language, including warnings that advanced AI systems could pose catastrophic or existential risks. That language is not a prediction that such an outcome will happen; it is the kind of disclosure a prospectus is required to make, and it reflects a debate Anthropic’s own chief executive, Dario Amodei, has raised publicly about the pace of AI development.
What is still blank, and what reports are only guessing at
Three things are not in the filing, and any article that states them as fact is getting ahead of the document.
First, the price range and share count are undetermined. Second, the listing date is not set. Third, the valuation is not set.
Press reports have floated valuations in the $1.5 trillion to $2 trillion range, and some reports have pointed to a possible mid-November debut, with other reports mentioning a roadshow in the week of 9 November. Those are reports, not filing terms, and they differ from each other. The filing itself leaves those fields blank, and an amended filing will be needed before pricing.
Goldman Sachs and Morgan Stanley are listed first among the underwriters, with a syndicate of more than 20 banks in total. Both Anthropic and some existing shareholders will sell shares; Anthropic will receive nothing from the shareholders’ sales.
On Monday, US shares rose, with the S&P 500 gaining 0.70% and the Nasdaq rising 1.10%, and chipmakers including Nvidia, AMD, Broadcom and Micron moved higher. That market move happened on the same day as the filing and alongside other AI funding news, including Meta’s plan to raise up to $95 billion, so it should not be read as a verdict on Anthropic alone.
What to watch next
The next concrete step will be an amended S-1 that fills in the share count, price range and timing. Until the registration becomes effective, no shares can be sold and no offers to buy can be accepted.
Watch three things in that amendment: whether the price range lands near the reported valuation range or well below it, how much of the offering is new Anthropic stock versus existing shareholders selling, and whether the governance structure changes in response to investor feedback.
Readers following today’s biggest stories can also catch up on the site’s Brewers Game 3 coverage, which landed in the same news cycle as the Anthropic filing.
Frequently asked questions about the Anthropic IPO
Is Anthropic a public company yet?
No. Anthropic filed a public S-1 on 26 October 2026, but the registration is not yet effective and no shares have been priced or sold.
What ticker will Anthropic use?
Anthropic has applied to list its Class A common stock on the Nasdaq Global Select Market under the symbol ANT.
Why did Anthropic report an $80.7 billion loss if its operating loss was only $693 million?
About $72.4 billion of the net loss was a non-cash fair-value charge on convertible notes, derivative liabilities and similar instruments, driven mainly by the rise in Anthropic’s own valuation. The operating loss is a separate, narrower measure, and neither figure on its own tells the full story.
How much is Anthropic worth in the IPO?
That has not been set. Press reports have floated $1.5 trillion to $2 trillion, but the filing leaves valuation, share count and price blank, so any single figure today is a report, not a confirmed term.
Who will control Anthropic after the IPO?
A founders’ entity will hold 50.1% of general voting power through a special Class F share, and Anthropic’s Long-Term Benefit Trust will generally elect a majority of the board. Public shareholders will therefore have limited practical control, by design.
