Context window
The maximum amount of text, measured in tokens, a model can consider in a single request.
The context window is the model's working memory for one request: the system prompt, the conversation history, any retrieved data, and the response all have to fit within it. Exceed it and the earliest content is dropped or the request is rejected.
A larger window lets you include more history or more retrieved documents, but it is not free — more tokens cost more and can dilute the model's attention, so relevant context still beats abundant context. Managing what goes into the window is a core part of building any non-trivial LLM application.