Post #2303170
2026-05-07 22:28 UTC
@toddsundsted@epiktistes.com @dangoodin@infosec.exchange I would argue they not have lowered the cost. I don't believe anyone is really being charged the actual cost of running these LLM beasts. What I want to know is what is the actual cost to run these models. At least tell us how many tokens were burned (or water evaporated) to get these results.
Replies (1)
-
@toddsundsted@epiktistes.com 2026-05-07 22:39
@EdBruce@infosec.exchange @dangoodin@infosec.exchange that's fair and i'm skeptical end users are paying the actual cost of inference right now, but i recently dropped $50K on a pen test. even at 100x the out of pocket costs, equivalent results from LLMs cost a fraction of that.