Anthropic Hits $16.8B in Q3 Revenue Before November IPO
Anthropic reportedly pulled in $16.8B last quarter and turned a profit, but slowing growth has some investors lowering expectations.
Business · Source: PYMNTS
What happened
Anthropic is gearing up for a November initial public offering with massive revenue numbers. The AI company reportedly told investors it generated 16.8 billion dollars in the third quarter. That is a 45 percent jump from the 11.6 billion dollars it made in the second quarter. It also represents a massive leap from the 4.8 billion dollars recorded in the first quarter. The company confirmed it was profitable on an adjusted basis for the second straight quarter.
Executives spent a full day Wednesday pitching investors at their San Francisco office. They argued that their safety-focused approach, fast growth, and better efficiency make them ready for the public markets. However, the growth rate is actually slowing down compared to earlier this year. This deceleration is falling below the expectations of some early backers.
This slowdown is causing a split in revenue projections. A Wall Street Journal report claims investors expect Anthropic to hit 110 billion dollars in annualized revenue by year-end. But research group SemiAnalysis projects a much lower 74 billion dollars. Joey Brookhart, an analyst at SemiAnalysis, noted that investors have been downgrading their projections over the last three months. Analysts point out that major clients like Meta are increasingly shifting to their own in-house AI tools.
Key facts
- $16.8 billion — Reported Q3 revenue for Anthropic
- $11.6 billion — Reported Q2 revenue for Anthropic
- $110 billion — Investor expectation for annualized revenue per WSJ
- $74 billion — SemiAnalysis projection for annualized revenue
Why it matters
The public market is about to test the real value of frontier AI models. Anthropic is proving that massive revenue and adjusted profitability are possible in the foundation model space. But the gap between 74 billion dollars and 110 billion dollars in projected annualized revenue shows that nobody knows how to price AI growth accurately. Builders need to watch this IPO closely. It will set the financial benchmark for every AI startup raising capital next year. If Anthropic struggles in the public market, private funding for AI wrappers and infrastructure will dry up fast.
The second-order effect is the shift in enterprise buying behavior. Major customers are starting to build and use their own AI tools instead of relying solely on closed models. If giants like Meta are pulling back to use internal tech, mid-market companies will eventually follow. The moat for API-based model providers is getting shallower. Companies are realizing that renting intelligence forever is too expensive. They want to own their infrastructure. This creates a massive opening for startups building tools that help enterprises transition away from expensive proprietary models.
For builders
Build around enterprise cost fatigue
Large customers are actively looking to replace expensive API calls with internal tools. If you build infrastructure that helps companies host and route to cheaper open models, enterprise buyers will pay you to lower their Anthropic bills. The big model providers will lose enterprise lock-in as a result.
Anchor valuations to actual revenue
The days of raising huge rounds on pure hype are closing. Anthropic is going public on the back of 16.8 billion dollars in quarterly revenue and adjusted profitability. Early-stage founders must show clear paths to monetization, or venture capitalists will walk away.
Target the fraud detection market
Card issuers are aggressively adopting AI to stop fraud without blocking legitimate sales. Founders who build real-time fraud scoring tools can sell directly to financial institutions. Legacy tech vendors will lose market share as issuers shorten their upgrade cycles to deploy AI faster.
My take
Hype gets you the meeting, but a 16.8 billion dollar quarter gets you the IPO. Anthropic is proving you can build safe AI and still print money. But with massive clients like Meta building their own tools, they better hope the public markets focus on today's profits instead of tomorrow's churn. I always tell founders to watch what enterprises do, not what they say. They are quietly looking for the exit door on expensive API dependencies.
Original reporting: PYMNTS. This is my rewrite and opinion.