Context window
Definition: The context window is the amount of text (in tokens) a model can consider at once — its 'working memory'.
A large window lets it analyze long documents or conversations. Beyond it, older information is dropped.
Frequently asked questions
What is a large context window?
The amount of text (in tokens) a model holds "in mind" at once. A large window — hundreds of thousands of tokens — lets it analyze long documents or conversations in one pass.
What happens beyond the window?
The oldest information falls out of scope and is "forgotten." That's why RAG helps: supply only the relevant passages instead of loading everything into context.