Product1 distinct publisher3 min readUpdated
Revenue of more than $11.5bn and a positive adjusted operating income line replace annualized run-rate maths with something a buyer can actually check before signing a multi-year deal.
The Product Desk · Product desk

Compiled by The Product DeskSomething wrong?How this is made
Anthropic has told prospective investors that second-quarter revenue exceeded $11.5bn, more than 14 times the $787mn it recorded in the same quarter of 2025, according to documents seen by Bloomberg [1]. It also recorded positive adjusted operating income, which TNW notes the leading AI labs have not been able to say [2].
The size of the number is less useful to an operator than its shape. The industry has spent two years asking to be valued on annualized run rates, a figure derived by taking current revenue and multiplying it out; Anthropic has now put a completed quarter and an operating income line in front of investors instead [3].
That distinction is checkable, and it checks. Anthropic disclosed a $47bn run rate in May [4]. Four quarters at $11.5bn annualizes to $46bn [5], roughly 2 percent below the May figure [6]. Run rates and completed quarters do not usually reconcile that closely, because a run rate is built from the best available month and a quarter is not. Sequentially, first-quarter revenue was $4.73bn, taking the first half to about $16.2bn [7], which makes Q2 about 2.4 times Q1 [8].
Comparisons need care. OpenAI's reported figure of more than $40bn is a run rate rather than a quarter's revenue, and Bloomberg notes the two may not be calculated the same way [9]. A buyer holding a vendor scorecard with both numbers in the same column has a broken scorecard.
The reason for the disclosure is a listing. Anthropic has filed confidentially and is working with Morgan Stanley, Goldman Sachs and JPMorgan, with backers telling the Financial Times they expect a $2trn valuation in October [10]. At $46bn annualized, that is about 43 times revenue [11]. An autumn debut would put Anthropic on the public market before OpenAI, whose valuation is under scrutiny from its own investors, and before DeepSeek, which is preparing to file [12]. The window is open: listings have raised $256.4bn this year excluding blank-cheque vehicles, the most since 2021 [13].
For anyone negotiating a two- or three-year commitment, the practical value of a completed quarter with a profit line is that it speaks to whether the vendor can fund the capacity it has promised without a rescue financing mid-contract. A run rate cannot answer that question, because it says nothing about cost.
The limits are real. The figures are preliminary and could be revised, and Anthropic declined to comment [14]. The income line is described as adjusted, and the reporting does not specify what the adjustment excludes [15]. Nothing here discloses gross margin, customer concentration, or the compute obligations sitting against that revenue, all of which matter more to a vendor's ability to honour a long contract than a headline growth multiple does.
Three things worth watching. Whether the eventual public prospectus restates the $11.5bn and defines what adjusted means [14][15]. Whether the third quarter holds the trajectory, since a single quarter that more than doubles [7] tells you about demand timing, not durability. And whether the October date and the $2trn expectation survive contact with public investors [10], because a repriced listing changes the capital available to the vendor sitting across the table.
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
Anthropic recorded positive adjusted operating income in the quarter, which TNW says is not something the leading AI labs have been able to say.
Anthropic has filed confidentially for a listing and is working with Morgan Stanley, Goldman Sachs and JPMorgan, with backers telling the Financial Times they expect a $2trn valuation in October.
The reported operating income figure is described as adjusted, and the reporting does not specify what the adjustment excludes.
Anthropic told prospective investors that second-quarter revenue exceeded $11.5bn, more than 14 times the $787mn it recorded in the same quarter of 2025, according to documents seen by Bloomberg.
The industry has asked to be valued on annualised run rates, a figure that takes current revenue and multiplies it out; Anthropic has now put an actual completed quarter and an operating income line in front of investors instead.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Single-publisher relay of preliminary, unaudited figures
One outlet in the cluster, itself relaying documents seen by Bloomberg plus an FT-sourced valuation expectation. The company declined to comment, the figures are explicitly preliminary and revisable, and the profitability line is 'adjusted' with no stated exclusions. The internal arithmetic does hold — annualising the quarter lands within about 2 percent of the previously disclosed run rate — which lifts the score above the floor.
Revenue scale implies substantial paid usage, no customer detail
More than $11.5bn of revenue in one quarter, up more than 14x year over year and roughly 2.4x sequentially, is direct evidence of large-scale paid commercial uptake rather than pilot interest. It is scored below the top band because the supplied reporting gives no customer counts, deployment examples, contract mix or concentration data, and the underlying figures are preliminary.
Modestly overstated: 'checkable' framing outruns unaudited inputs
The reporting is comparatively disciplined — it flags preliminary status, the adjusted metric, and the non-like-for-like OpenAI comparison. But the framing that a completed quarter is something a buyer can actually check overstates what preliminary, unaudited, undefined-adjustment figures relayed second-hand can support, and the $2trn / roughly 43x revenue expectation is carried forward as context rather than interrogated. Net tilt is mildly positive rather than severe.
Pre-listing figures shown to prospective investors by an unnamed-document route
The numbers were prepared for prospective investors ahead of a confidential listing being run by three underwriting banks, with backers already circulating a $2trn expectation and timing framed as beating OpenAI and DeepSeek to market. That is a strong incentive structure to present favourable, favourably adjusted preliminary figures, and the company would not comment on the record.
Low-to-moderate: coherent numbers, thin sourcing
Confidence is limited by a one-publisher cluster relaying documents it did not obtain, preliminary figures, an undefined adjustment and no company comment. It is not lower because the specific figures are internally consistent, the source names its own upstream reporting, and the article volunteers its comparability and revision caveats.
product
Anthropic's answer to a 2 percent stake: hand the founders extra votes1 distinct publisher
product
Anthropic wants a bigger raise than SpaceX and will not say what it is worth2 distinct publishers
invest
The first AI IPO writes the comp sheet, and Anthropic is holding the pen1 distinct publisher
build
The $559M-versus-$12.3B quarter matters more than the $65B run rate4 distinct publishers
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 15, 2026