Post

Whenever I read discourse on AI energy/water use that focuses on the "median query," I can't help but feel misled. Coding agents like Claude Code send hundreds of longer-than-median queries every session, and I run dozens of sessions a day. On my blog: www.simonpcouch.com/blog/2026-01...

Alex Bradbury

@asbradbury.org

20 Jan

Nice post! You might also be interested in my attempt to get some figures on inference energy usage, but coming from the perspective of the concrete data available in the InferenceMAX benchmarks. muxup.com/2026q1/per-q... Many many provisos and limitations of course

muxup.comPer-query energy consumption of LLMsCan we reasonably use the InferenceMAX benchmark dataset to get a Wh per query figure?

Simon P. Couch

@simonpcouch.com

I very much enjoyed reading this post! It's cited in the footnotes as one reason not to trust my use of the 5:1 IO based on Anthropic's pricing. :)

21/1/26, 12:04 am0 reposts3 likes

No replies yet.