Guide
Glossary
Context window
A context window is the maximum amount of text, measured in tokens, that an AI model can consider at one time. It includes everything in the request: the instructions, any attached documents, the conversation so far and the model's own reply. Anything beyond the limit is cut off, summarised or never seen. As at September 2026, several leading models accept about a million tokens, roughly 750,000 words, although chat products may allow less. A bigger window is not automatically better: Anthropic's documentation notes that accuracy and recall degrade as the token count grows, so sending only the relevant material usually gives better answers.
Also called context length, context limit, token limit
Last updated:
Example
In a law firm
A litigation team asks an AI tool to summarise a 5,000-page discovery bundle. The bundle is larger than the tool's context window, so the tool either truncates it or picks the passages it judges relevant, and the summary can silently miss documents. The team instead processes the bundle in batches, with each summary citing page references a paralegal can check.
Guides that explain it in context
Secure by design. Set up correctly. Fully managed.
Talk to us before you commit to anything
Start with a free 45-minute discovery call. We look at your systems and priorities, then recommend a first step with a fixed scope, or tell you if we are not the right fit.