The PracticeThe PracticeThe PracticeOperating infrastructure

The Practice · Glossary

Inference

AI term; checked November 9, 2026

Inference is a trained model at work: taking an input and producing an output. Training happens once, at the lab; inference happens every time anyone sends a message, and it is what usage-based pricing charges for.

A tool that sends your whole file with every request costs more than one that sends only what the step needs. Ask any tool you pay for by usage how it decides what it sends. See Claude vs ChatGPT for Real Estate Agents.

Educational, not professional advice. Nothing on this page is affiliated with or endorsed by its publisher.