Context window
What is Context window?
The amount of text a model can hold in mind at once, measured in tokens. Think of it as working memory: the current conversation plus any documents you paste all have to fit inside it. Once you go past the limit, the model starts losing track of what came earlier.
Why it matters
The context window is how much text a model can hold in mind at one time, counted in tokens. Think of it as working memory. The instruction, the current conversation and any documents you paste all have to fit inside it at once. When you go past the limit the model starts losing track of what came earlier, so a long chat can suddenly forget a detail you gave it at the start. For anyone building with AI, the window size sets a hard ceiling on how much material the model can consider in a single answer.
In practice
A team feeding a model a 50-page contract has to check it fits the window, or the model only “sees” part of the document and answers as if the rest does not exist. This is a common cause of confidently wrong answers on long inputs. Larger windows ease the problem but cost more per call, since you pay for every token in that memory. The real skill is deciding what actually needs to be in the window and leaving out what does not.
