Anthropic's Q3 revenue hits $16.8B but growth slows ahead of planned November IPO

Anthropic told potential investors at a private meeting in San Francisco that its third-quarter revenue reached $16.8 billion, according to multiple US media reports. That figure represents roughly a 45 percent increase over the second quarter and is more than 3.6 times the company's total annual revenue for 2025. The company is preparing for an initial public offering that The Information says is planned for November, before Thanksgiving.
The headline number is large, but the growth rate is the part that appears to have unsettled the room. In the second quarter, revenue rose more than 140 percent from the first quarter, driven by surging demand for the developer tool Claude Code. That pace has not held. A summer slowdown is one likely factor, since people typically work less during those months and consume fewer tokens. Competitive pressure is another: OpenAI launched a strong rival product called Astra in September, and Anthropic's answer, Opus 5.5, only arrived late that month.
The company also gave investors a profitability picture. According to The Information, Anthropic reported an adjusted operating margin of 11 percent for the year through the third quarter. That measure excludes major costs such as stock-based compensation. The Financial Times reported it was the second consecutive quarter of adjusted profitability, and the company is cash-flow positive: during the third quarter it generated $1.70 in revenue for every dollar spent on compute. For an industry that has spent years arguing about whether large models can ever pay for themselves, that ratio is the most concrete signal yet that inference on a trained model can be a viable business rather than a subsidy contest.
Caveats remain substantial. Training runs and staff costs are steep, and open-weight alternatives keep improving, which threatens the pricing power of proprietary models like Claude. Adjusted profitability also is not the same as profitability under standard accounting, since stock-based compensation is a real cost to existing shareholders through dilution even when it does not appear in operating margin.
Anthropic also disclosed that an estimated 5 to 7 percent of its revenue comes from distillation, a practice in which other companies query its models to extract information about how they behave and then train competing systems on the results. Anthropic has repeatedly accused Chinese AI labs of copying its models this way. The company recently rolled out countermeasures that it describes as effective, which raises an awkward question about why it lobbied against distillation for so long if it had the technical means to stop it. Anthropic told investors that distillation revenue will likely decline now and weigh on the current quarter's results. Management's stated long-run case is that blocking the practice makes it harder for rival model makers to match Claude's quality.
The valuation conversation is where the reported numbers are landing hardest. Anthropic's private valuation reached $965 billion in May, up from under $20 billion two years earlier, and some bankers had floated a $2 trillion figure for the offering. Several potential investors were reportedly disappointed by the revenue disclosure and now expect Anthropic's banking advisors to cut valuation expectations to possibly below $1.5 trillion. The company did not discuss its September annualized revenue rate or provide a full-year outlook; in July that rate was reportedly $65 billion.
The market backdrop is not friendly. Inflation remains high, borrowing costs are rising, public sentiment toward AI is more negative than positive, and unresolved cybersecurity risks represent an operational threat to AI providers. OpenAI, Anthropic, Google and others have been locked in a price war for months while open models close the performance gap. Countering that, the case for long-run demand is straightforward: as model capabilities improve, they work their way into more everyday office tasks, and usage tends to grow rather than shrink.
What the meeting did not settle is how public investors will price a company whose revenue is enormous but whose quarter-over-quarter growth curve has visibly bent. A 45 percent sequential gain would be remarkable for almost any software business; for an AI lab that recently posted 140 percent, it reads as deceleration. Whether the IPO prices near $1.5 trillion or higher will depend on whether buyers treat the margin and cash-flow numbers as proof of a durable business or the growth slowdown as the more important signal.
Why it matters: This is the first hard look at the financials of a frontier AI lab heading to public markets, and it sets a reference point for how investors value the whole sector. Developers and product teams should watch the distillation crackdown, since cutting off model-extraction revenue could change the availability and pricing of API access that some downstream tools depend on. If the offering prices below expectations, that is likely to pressure valuations and spending plans across AI infrastructure and tooling, though the source gives no indication of how Anthropic itself would respond.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
Anthropic hit $16.8 billion in third-quarter revenue, a 45 percent gain that still falls far short of Q2's 140 percent surge. With a November IPO approaching, some investors now expect the company's valuation to drop below $1.5 trillion. Slowing growth and a tough market are cooling expectations for the AI boom's biggest Wall Street test yet. The article Anthropic's Q3 revenue hits $16.8B but growth slows ahead of planned November IPO appeared first on The Decoder .