SAN FRANCISCO THURSDAY, OCTOBER 15, 2026
By almost any normal standard, the numbers Anthropic shared with investors on Wednesday were spectacular. The Financial Times reported, according to people familiar with the matter, that the Claude maker's revenue reached $16.8 billion in the quarter that ended September 30 — up roughly 50 percent from the previous quarter. Reuters, citing an investor who attended the pre-IPO meeting in San Francisco, put the growth at 45 percent and the prior quarter at $11.6 billion, with first-quarter revenue at $4.7 billion. And yet, within hours, the story in the markets was not the size of the number. It was the direction of the growth rate.
A quick but important note on sourcing before we go further: Anthropic is still a private company. These figures were shared at an invite-only investor day and reached the public through reporting by the Financial Times, Reuters, Bloomberg and others citing people familiar with the matter or investors who attended. They have not been published as audited financial statements, Anthropic did not immediately respond to requests for comment from several outlets, and the reports differ slightly — 45 percent growth in one account, roughly 50 percent in another, a second quarter of $11.5 billion here and $11.6 billion there. Treat the shape as well-sourced reporting, not as a filing.
The headline numbers, as reported
Here is what the reporting consistently says. Third-quarter revenue: $16.8 billion. Growth from the second quarter: roughly half again, after growth of roughly 140 percent between the first and second quarters. Profitability: the company told investors it was profitable on an adjusted basis for the past two quarters and cash-flow positive for the past quarter, with an adjusted operating margin reported at 11 percent. Reuters also reported that the company is targeting a valuation of over $2 trillion at its IPO, which is expected to come after the November U.S. midterm elections, and that its annualized revenue run rate topped $65 billion earlier this year in prior reporting.
Adjusted figures deserve their usual asterisk: they exclude some costs, and for a company spending at Anthropic's scale, the excluded costs are enormous. Multiple reports noted a net loss of $42 billion last year on a full accounting basis, and Reuters reported last month that Anthropic has committed hundreds of billions of dollars in future spending under long-term agreements for cloud services, computing and infrastructure — with one report putting planned spending on chips and infrastructure at over $500 billion across the next decade. Both things can be true at once: a genuinely profitable core operation on an adjusted basis, and a capital plan of almost incomprehensible size.
So why did the AI trade wobble?
Because at a $2 trillion valuation, you are not pricing a good quarter — you are pricing a trajectory, and trajectories are judged on their slope. MarketWatch reported that the deceleration from roughly 140 percent quarterly growth to roughly 50 percent was read by investors as a red flag for the wider AI trade, noting that a Jefferies analyst had written before the investor day that Wall Street was likely anticipating an annual revenue run rate of at least $80 billion, with hopes of surpassing $105 billion by year-end. Barron's reported the same dynamic from the market side: the numbers were strong, the expectations were stronger, and AI-linked stocks — chipmakers and storage names among them — fell again in early trading on Thursday.
There is something faintly absurd about a company growing revenue by half in a single quarter being treated as a disappointment, and it is worth saying so plainly. Almost every business on Earth would frame $4.7 billion to $16.8 billion across three quarters of a single year as a triumph. But public markets do not grade on a curve of sanity; they grade against the price already paid. At the valuations now attached to the frontier AI labs, even excellent growth can be bad news if it is slower than the growth the price assumed.
The IPO stakes
That is what makes Anthropic's debut such a consequential test. Reuters reported that the IPO is expected to precede a public listing by rival OpenAI, which means Anthropic will be the first of the frontier labs to have its finances picked apart in public, quarter after quarter, by investors who can sell as easily as buy. Its bankers will be asking the market to believe that breakneck revenue growth can eventually outrun the enormous cost of building and running frontier AI systems — the chips, the data centers, the long-term cloud contracts. Wednesday's investor day was the opening argument in that case, and the market's response suggests the jury is genuinely undecided.
The bulls and bears can both find their paragraph in these numbers. Bulls: $16.8 billion in a quarter, adjusted profitability, positive cash flow, and a product — Claude — that has become a default choice for a large share of enterprise AI spending. Bears: decelerating growth, a $42 billion net loss on the books last year, and spending commitments measured in the hundreds of billions against revenue that, however fast it grows, is not yet in that league. Both paragraphs are accurate. That is precisely why this IPO matters far beyond one company: it is the first real public referendum on what frontier AI economics are actually worth.
What to watch
The next milestones are mechanical but telling: a public filing, which will replace investor-day reporting with audited disclosure; the final pricing, which will show whether the $2 trillion ambition survives contact with public investors; and the first reported quarter as a public company, when the growth-rate question gets asked again with real consequences. Until the filing lands, every number in this story carries the caveat at the top of it — reported, sourced and plausible, but not yet audited. We will update our coverage when the paperwork makes it official.
More from Previously
- Hot inflation sinks tech stocks for a second day even as TSMC posts a record quarter
- BofA's $1.2 billion charge, Morgan Stanley's beat — and Thursday's data gauntlet
