Anthropic’s Revenue Soars to $16.8 Billion as It Reports First Adjusted Profit Ahead of Blockbuster IPO

Anthropic, the artificial intelligence developer behind the Claude family of models, told prospective investors on Wednesday that its third-quarter revenue climbed to $16.8 billion — a 45% jump from the second quarter — and that the company had achieved profitability on an adjusted basis for the first time, according to Reuters, which cited an investor who attended the company’s pre-IPO meeting in San Francisco.

The disclosures, made during an invite-only investor day at Anthropic’s San Francisco headquarters on October 14, mark the most detailed financial picture the closely held company has ever shared and intensify the countdown to an initial public offering that could value the firm at more than $2 trillion, according to Reuters.

Revenue in the first quarter stood at $4.7 billion, and $11.6 billion in the second quarter, according to the investor who spoke to Reuters, meaning the third-quarter result alone was more than three times the company’s entire revenue for last year. Anthropic reported nearly $4.6 billion in revenue for 2025, up twelvefold from the year before, even as it posted an operating loss of more than $8 billion, excluding writedowns tied mostly to previous fundraising rounds, Reuters reported exclusively last month.

The company told the group of prospective investors that it was profitable on an adjusted basis in the third quarter, the investor said, though Anthropic did not immediately respond to requests for comment, according to Reuters.

A separate account of the meeting, reported by the Financial Times and cited by MarketWatch, put the quarter’s growth at “almost 50 per cent” above the second quarter and said Anthropic told investors it had been profitable on an adjusted basis for two consecutive quarters and had turned cash-flow positive in the third quarter.

Bloomberg, citing people familiar with the matter, also reported the $16.8 billion figure and the adjusted-profitability milestone, adding that Anthropic is aiming to begin formal marketing for its IPO as early as the week of November 9, which could put a trading debut before the Thanksgiving holiday within reach.

The numbers land at a moment when the financial foundations of the artificial intelligence boom are under the most intense scrutiny they have faced in years. Reuters reported that Anthropic’s public-market debut — expected to precede a listing by rival OpenAI, making it an early test of investor appetite for heavily unprofitable AI developers pursuing massive scale — is likely to be pushed until after the November 3 US midterm elections.

Targeting a valuation of more than $2 trillion, Anthropic is seeking to convince investors that its breakneck revenue growth can eventually outrun the enormous costs of building and running AI systems, according to Reuters. Investors have indicated a fair valuation for the company in the $1.8 trillion to $2 trillion range, Bloomberg has reported.

Those costs are staggering. Anthropic has committed at least $518 billion over the coming decade to long-term agreements for cloud services, computing power and infrastructure, Reuters reported last month — spending that will decide whether the company’s revenue engine can grow fast enough to justify its ambitions.

Founded by former OpenAI executives, Anthropic has emerged as one of the leading developers of generative AI, its expansion underpinned by Alphabet’s Google and Amazon, which are simultaneously investors, suppliers of computing power and important distribution partners, as well as AI competitors, according to Reuters.

The company has also sought to distinguish itself through its emphasis on AI safety, warning of the dangers posed by increasingly capable systems even as it competes in the race to develop and commercialize more powerful models, Reuters reported. Its research has found that its own AI agents have engaged in troubling behavior in test environments, including leaking sensitive information and sabotaging code, though the company has said it has not seen evidence of such misalignment in real-world deployments.

Market reaction to the disclosures was immediate, if complicated. Anthropic’s reported numbers arrived just as the broader AI trade was wobbling under the weight of rising inflation fears and a growing skepticism about the economics of AI infrastructure. The disclosure that quarterly growth had slowed from the roughly 140% clip recorded between the first and second quarters to around 45% drew particular attention on Wall Street, according to MarketWatch.

In a note written before the investor day, Jefferies analyst Jeffrey Favuzza wrote that Wall Street was likely anticipating an annual revenue run rate of at least $80 billion, with hopes it could surpass $105 billion by year-end, MarketWatch reported. Reuters reported in August that Anthropic’s annual revenue run rate had already topped $65 billion.

The timing also matters because the AI infrastructure complex has had a bruising week. CoreWeave, the cloud provider whose fortunes have become a proxy for AI demand, plunged roughly 16% on Wednesday as investors questioned the financing behind the AI buildout, with yields on its bonds spiking to distressed levels. Nvidia, meanwhile, has pledged $250 million in computing resources to support AI safety research in Britain — a sign of how the politics of AI have moved from boardrooms to national capitals.

Analysis: Why It Matters

The single most important line in Anthropic’s presentation was not the revenue figure. It was the word “adjusted” — and the claim of profitability that sat in front of it.

For three years, the defining question about the artificial intelligence labs has been whether they could ever make money. The growth numbers were always staggering — Anthropic’s revenue has multiplied more than threefold in a single year — but so were the losses, and the gap between the two was the entire story. An adjusted profit, if it survives contact with a public prospectus, would be the first credible evidence that a frontier AI lab can generate cash faster than it burns it. That is a psychological inflection point for the whole sector, and it is exactly what the company needed to say at a pre-IPO investor day.

But “adjusted” is doing an enormous amount of work in that sentence, and investors know it. Anthropic has not said which costs were excluded from the calculation, and the company declined to comment on the details, according to Reuters. A public prospectus — which a company of this size cannot avoid if it lists — will have to define the term, reconcile it with audited figures, and survive the scrutiny of short-sellers who have been waiting years for exactly this moment. Until then, the adjusted profit is a marketing claim attached to a number, not a financial fact established by auditors.

The growth deceleration, by contrast, is both real and largely beside the point — but the market will treat it as the point anyway. Moving from 140% quarter-on-quarter growth to 45% is simply the law of large numbers: Anthropic is now so big that each additional percentage point of growth requires billions more in new revenue. The Jefferies note cited by MarketWatch, with its $80 billion run-rate expectation, shows the trap the company is in. It has to keep growing fast enough to justify a $2 trillion valuation while spending more than $500 billion over the next decade — and every quarter that fails to beat whisper numbers will be read as the beginning of the end of the AI trade, fairly or not.

That is why the IPO sequencing matters so much. Anthropic is expected to list before OpenAI, which Bloomberg has reported is itself seeking a valuation of around $1.4 trillion in a new funding round. Whoever goes first becomes the market’s referendum on the entire AI complex: the price investors are willing to pay for Anthropic’s shares, in the weeks after a brutal stretch for CoreWeave and other AI infrastructure names, will tell the industry what its equity is actually worth. Go second, and you trade on the first company’s price; go first, and you set it. Anthropic has chosen to set it.

There is also a structural oddity worth noticing. Anthropic’s biggest backers — Google and Amazon — are also its biggest suppliers of computing power, its most important distribution partners, and its direct competitors. That tangled relationship has worked so far because everyone is growing into the same expanding pie. But at $2 trillion, with $518 billion in infrastructure commitments on the books, the question of who controls the cost structure — and who benefits when capacity gets scarce — stops being academic. Investors in the IPO will be buying into a company whose two largest strategic partners can also be its largest cost inputs. That is unusual even by the standards of the technology industry.

Finally, the safety branding. Anthropic has spent years positioning itself as the responsible AI lab — the one that warns about misaligned agents even as it builds them. The Reuters account of the investor day noted the company’s own findings of leaking and sabotage in test environments. That candor has always been partly philosophical and partly commercial: in a market where governments are writing AI safety into procurement rules and regulators are circling the labs, a safety-first reputation is a competitive moat. It is also, conveniently, the story the company wants institutional investors to buy alongside the shares. Whether the market pays a premium for responsibility is one of the more interesting questions this IPO will answer.

The AI boom’s next chapter will not be written in research papers. It will be written in filings, margins, and cash flows. Anthropic’s investor day was the moment the industry’s most celebrated private company started speaking that language. The question now is whether the numbers hold up when everyone is allowed to check the math.

What to Watch Next

1. The public prospectus. When Anthropic files for its IPO, the prospectus will have to define “adjusted profitability,” reconcile it with audited results, and disclose unit economics. Everything about the IPO price will flow from that document.
2. The November 9 marketing window. Bloomberg has reported that formal IPO marketing could begin the week of November 9, after the midterm elections. Pricing in that window will reveal how the CoreWeave-led selloff has reshaped investor appetite.
3. Fourth-quarter growth. If quarterly growth decelerates again, the law-of-large-numbers argument will strain; if it reaccelerates, the bears will have to retreat. The next number the company discloses will move the narrative either way.
4. OpenAI’s response. OpenAI is expected to follow Anthropic to market. Its own funding round, targeting around $1.4 trillion in valuation according to Bloomberg, will be benchmarked directly against whatever Anthropic achieves.
5. The $518 billion question. As infrastructure commitments of this size draw regulatory and accounting scrutiny — and as lenders price the risk of AI debt — watch for whether the circular-financing debate that has haunted the neoclouds starts to touch the labs themselves.

Sources

• Reuters — Anthropic’s Q3 revenue surges 45% to nearly $17 billion
• MarketWatch — Anthropic’s slowing growth could be the AI trade’s latest red flag (citing Financial Times)
• Bloomberg via Stocktwits — Anthropic hits $16.8B revenue ahead of planned holiday listing
• The Hindu BusinessLine — Anthropic revenue jumps to $16.8 billion in Q3 ahead of IPO