← Concept Index

Model foundations

Knowledge cutoff

Also called: training cutoff, up to date

DEFINITION

The knowledge cutoff is the point in time after which a model saw no training data. Events, prices, and releases after that date are outside what it learned.

WHY IT MATTERS

A model will still answer confidently about things after its cutoff — by guessing. For anything time-sensitive, you need retrieval, a web tool, or a human check, not the model's memory.

COMMONLY CONFUSED WITH

The context window. You can paste today's news into the window even if it is after the cutoff; the cutoff is only about what was learned during training.