Skip to content
BENIBIX Learn

Knowledge Cutoff

AI💬 LLMsLanguage Model Limits

Also known as Training Cutoff, Training Data Cutoff

The date a model's training text ends. It knows little after that.

Think of it like

Think of it like a letter sealed in a time box. It only knows what was true on the day it was sealed.

Example

Emma asks a chat model about a phone launched last month. The model's cutoff was a year ago, so it guesses, or says it does not know.

Why it matters

Anything new, from prices to laws, may be missing from the model's answer. Or it may be out of date.

Definition

A knowledge cutoff is the date after which a model's training text stops. The model knows little or nothing about events after it. A model is often released months after its cutoff and used for a year or more beyond that. So it may not know about new products, prices, laws or news. Tools that let the model search the web or read your files fill that gap.
The model learned from text up to the cutoff. Everything after it is a blank.

Common mistake

Thinking the model knows today's date or news. Unless a tool gives it the date or a search, it does not.

Key takeaways

  1. 1
    A fixed date. Training text stops there.
  2. 2
    Blind after it. New events are unknown.
  3. 3
    Search fills the gap. Fresh sources at question time.

Compare

Learn it properly