Context Window
How much text a model can hold in view at once.
Technical Term
Full explanation
A context window is the amount of text an AI model can consider at one time when generating a response.
It is measured in tokens, and it is finite, ranging from a few thousand to several hundred thousand tokens depending on the model. Everything the model can "see" when generating a given answer, the question, any retrieved documents, prior conversation turns, has to fit inside that budget together.
Because the window is limited, a system answering from a website is almost never reading the whole site at once. It is more often working from a handful of retrieved pages or extracts chosen as most relevant, which is why a page's most important claim needs to be stated plainly, early, and in a self-contained way, rather than requiring the reader to piece it together from context spread across a long page.
Why this matters
It is finite, so systems work from selected extracts rather than whole websites. Content that states its point plainly and early travels better.
Example
A model with a 128,000-token context window can hold roughly a long book's worth of text at once, but a large site with hundreds of pages still will not fit. In practice, the system retrieves and feeds it a handful of the most relevant extracts, so a claim buried deep in a long document with no self-contained summary may never make it into the window at all.
Where this sits
Context Window belongs to AI & Search Infrastructure, one of the seven layers of the Aingworth Glossary. The technical vocabulary behind the interfaces. Useful for understanding how the systems work, without becoming the center of the conversation.
Related terms
Ready to see where your brand stands?
Start with the free diagnostic to see which condition of the Recognition Triangle is leaking, or read the frameworks these terms belong to.