Context Window
The maximum amount of text — prompt, history, documents and reply — a model can weigh at once, counted in tokens.
Definition
The context window is a model's working memory: the prompt, the conversation so far, any documents you supplied, and the answer being written must all fit inside it, and its size is counted in tokens. Think of it like a desk that holds only so many pages at once — pile on a book longer than the desk, and some pages have to come off before the model can read the rest.
Why it matters
When a document or conversation is bigger than the context window, something has to be left out — which is why an assistant can forget what you said earlier or miss part of a long file. AskClear works around this in Facts mode by selecting the passages most relevant to your question instead of blindly cramming in a whole report, so the answer stays grounded in the parts that actually matter.