.arunim.fyiIn the backseat of an Uber with my friend Richard a few days ago, I was trying to wrap my head around the fact that Anthropic’s revenue had been growing at a rate of 80x/year for the last few months. Bonkers! Insane! Unreal! I had also written down not one month earlier my “bullish” prediction that the 10x revenue would continue for 2026, if anything I’d hedge to 8x/year. Instead, the revenue growth rate increased! Not only that, it increased to 80x/year. Applying some easy math from their ARR of 800B ARR at the end of the year if this keeps up. Which is… incredibly large! I went to find reference points for other companies with this much revenue, and turns out they literally don’t exist—Anthropic would become the largest company (in terms of ARR) within 7 months if this continues.
Jesse Richardson then posted basically the same rant to his excellent Substack.
at its current rate of exponential growth, Anthropic ARR will hit 100% of current global GDP, i.e. around $126 trillion dollars, in December 2027. This would probably be enough to give them a valuation in excess of a quadrillion dollars. That’s an insane fact about the world. What makes it even more insane is that if you look at 2025, Anthropic’s revenue was growing at an absurdly fast rate for that year, fast enough to reach 100% of global GDP by May 2030. You could reasonably look at that and say sure, their revenue growth is impressive, but the line can’t keep going up, it has to taper off eventually. No way they’ll hit that number by mid-2030. You’d probably have felt pretty confident in making that claim, I know a lot of people did. But what did the line do? It bent upwards. Revenue started growing at a cool 80x per year, instead of 10x per year.
At some point in this process I realized that I still don’t have a good sense for how exponentials work. Like, what do you mean that they went from 10B to 45B in 4 months and will end up at 800B in 8 more months?? Extremely unintuitive. I still don’t think I’ve wrapped my head around it.
Some reference points:
- $10B annual revenue is less than Wegmans, an American supermarket chain which has 114 stores along the East Coast, or Wawa, or QuikTrip, or Trader Joe’s.
- $100B annual revenue is comparable to Nestlé or Alibaba
- and $1T annual revenue has never been seen before.
Coping mechanisms
They’re growing in a limited market
I suspect most of this growth is in the coding / software market, rather than elsewhere, and the software market is a) especially quick to adopt, b) limited in size, and c)
They are mostly growing by taking competitors’ market share
And so, if we’re trying to track the overall revenue of AI companies, then Anthropic’s rapid growth in particular isn’t as indicative of future performance.
See, e.g. this Ramp data
Compute / data centers will constrain them
What’s the ratio of revenue to compute? How’s it been changing?
Anthropic 2024 revenue: 1B Ratio: 0.38
OpenAI 2024 revenue: 1.8B Ratio: 2.06
Anthropic 2025 revenue: 2.7B Ratio: 1.67
OpenAI 2025 revenue: 8.0B Ratio: 1.625
How much will compute grow over time? certainly not 80x/year or even 10x/year.
We can look at electricity as well.
Open-weight models / cost reduction in the future
Costs will come down as open-weight or generally smaller, cheaper models are able to do the bulk of this kind of software work.
Jevons paradox
on the other hand, cost down ⇒ revenue up is somewhat well-documented.
The adopters are particularly excited to token-max right now
Google drive link: 1M context Claude leaderboards This is not normal!