Context Window
Also known as: Context length
The context window is the amount of text an AI model can take into account at once for a single request.
A language model doesn't work with unlimited memory, but with a fixed amount of text it can read at once for a given request — the context window. Everything that fits, whether that's the conversation so far, a pasted-in document, or the question itself, feeds into the answer. Anything that doesn't fit is simply unknown to the model for that request.
In practice, the context window shows up in how much text can be pasted into a chat before a model starts ignoring or forgetting earlier parts of the conversation, or whether a longer document can be processed in one piece at all. Models differ considerably in how large this window is.
One limiting case is assuming a larger context window automatically means better answers. Even within a large window, a model can overlook or under-weight details buried in the middle of a long text, and for a stable, growing set of documents, a searchable RAG system is usually more practical than reloading everything into the context window on every request.
What it means in practice
For a business, the context window matters when checking whether a chat tool can process a whole contract or a long set of minutes in one go, or whether the text needs to be split up first. For larger, ongoing document collections, a RAG connection is generally the more practical approach than aiming for the largest possible context window.
Is this handled properly at your site?
We look at how it actually stands with you — and say honestly whether anything needs doing.
Related terms
All termsA term from your quote missing here?
Send us the passage you do not follow. We will explain it — with no obligation to order anything.