# Anthropic's S-1 shows inference revenue per megawatt nearly quadrupling in a year

> Original title: Anthropic’s S-1 shows inference revenue per megawatt nearly quadrupling in a year.

- Source: SemiAnalysis
- Published: 2026-10-27T20:15:23.000Z
- Bad Influence: https://badinfluence.news/items/wg937ozvu66s2eguxzz9es31h
- Original: https://x.com/SemiAnalysis_/status/2115175550465507710

## Summary

SemiAnalysis reports that Anthropic's S-1 filing shows inference revenue per megawatt rising from $16 million last year to over $60 million currently, across its entire fleet. The filing says revenue grew 1,368% year over year for the three months ended June 30, while compute and infrastructure expenses grew 300%. The post also notes that Anthropic runs TPU, Trainium, NVIDIA GPU, and AMD GPU hardware, and that software matters as much as the hardware.

## Why it matters

The post quotes figures from Anthropic's filing on inference revenue per megawatt, a measure of how efficiently compute turns into revenue across its fleet.

## Post

Anthropic’s S-1 shows inference revenue per megawatt nearly quadrupling in a year.

“Inference revenue per megawatt went from sixteen million last year to sixty plus currently, across the entire fleet. AI model businesses scale extremely well.”

“Revenue grew 1,368% year over year for the three months ended June 30, compared to compute and infrastructure expenses that grew 300% during the same period.”

“How much of that is Anthropic being geniuses versus market-wide efficiency? We’re seeing GB be 50x better than Hopper.”

“They have TPU, Trainium, NVIDIA GPU and AMD GPU all online. New hardware is always more performant, but you need the software to use it. Who’s got the best software? Probably the company with the top engineers and the best model.”

[Video](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fext_tw_video_thumb%2F2115175434115571713%2Fpu%2Fimg%2FLuE8x63lQN2MMVtu.jpg&mode=thumb&exp=1793404800&sig=63a42c5e81b3db6a)