Anthropic’s pitch to investors began with a number built to cut through the AI-bubble anxiety: $16.8 billion in third-quarter revenue. The Claude maker said Wednesday that it had also reached profitability on an adjusted basis, according to reporting from Business Insider.¹

The figure followed reported revenue of $4.8 billion in the first quarter and roughly $11.6 billion in the second, putting third-quarter growth at about 45%. That is still explosive by conventional standards, powered by strong adoption of Claude Code and enterprise demand for its models. But it is a marked deceleration from the prior quarter, when sales more than doubled.

Anthropic is using the results to build support for a potential November listing that could value it near $2 trillion — potentially the largest IPO on record. Its more upbeat backers argue the company has offered a rare answer to Wall Street’s central AI question: can the immense cost of building advanced models translate into a real business? Axios reported that Anthropic told investors it was cash-flow positive in the quarter, calling it evidence that frontier AI could be sustainable even as major technology companies struggle to contain infrastructure spending.²

The bullish case rests on more than headline revenue. Claude Code’s annualized revenue was estimated at $6 billion to $7 billion by late summer, while Anthropic’s premium pricing and enterprise customers have helped support margins, Axios reported.²

Yet the company’s disclosures leave important gaps. It is unclear whether the $16.8 billion is gross or net revenue, or precisely how Anthropic defines adjusted profitability. PitchBook analyst Harrison Rolfes warned that the company is “still growing, but not growing as fast,” and said the slowdown could “shave” its prospective IPO valuation.¹

That tension will define the road ahead. Anthropic must fund vast training and computing commitments while competing against OpenAI, deep-pocketed rivals and cheaper open-weight models. The quarter gives it a powerful IPO narrative; sustaining it will require proving that cash generation survives the AI arms race.