Data/ML/AIGeneralHigh signalEmergingAround since 2023
Cost Per Token
Also written as Inference Cost, Token Economics
The unit economics of running a model — cost scales with text in and out, so a feature that looks cheap in testing can be ruinous at scale.
Think of it like
A phone plan billed by the word: fine until the whole company starts talking.
Junior or senior?
A strong product-sense signal in AI engineering.
Senior sounds like
Has changed a design — shorter context, caching, a smaller model — specifically to control cost.
Ask them
“What did an AI feature cost to run, and what did you change to bring it down?”