The Basic Definition
A context window is the maximum amount of text — measured in tokens, not words or characters — that a model can consider at once within a single conversation or request. Anything beyond that limit either gets cut off or must be summarized and dropped to make room.
Why Size Matters
A larger context window lets a model handle longer documents, longer conversations without losing earlier context, and tasks that require holding a large amount of information in view at once (like analyzing an entire codebase or a lengthy contract). Models with small context windows can lose track of earlier instructions or details in a long conversation.
The "Lost in the Middle" Problem
A large context window doesn't guarantee a model uses every part of it equally well — research has repeatedly found that models tend to pay more attention to information at the very start or end of a long context, sometimes underweighting details buried in the middle. A bigger number on a spec sheet doesn't automatically mean uniformly reliable use of all that context.
Practical Implications
For most everyday tasks, context window size matters less than people assume; it becomes a real factor mainly for long-document analysis, extended multi-turn conversations, or large codebases. See our Context Window Calculator for a rough way to estimate whether your specific content fits within a given model's limit.
Related Pages
Frequently Asked
Is a bigger context window always better?
Generally more capable for long-content tasks, but not a guarantee of uniformly reliable use of that entire context — see the 'lost in the middle' section above.
What happens if my input exceeds the context window?
Depending on the platform, it may be truncated, rejected with an error, or the system may attempt to summarize earlier content to make room — behavior varies by provider and product.
Is context window the same as memory across sessions?
No — context window is the limit within a single conversation or request; persistent memory across separate sessions is a different feature some platforms offer separately.
How do I estimate if my document fits a model's context window?
See our Context Window Calculator and Token Calculator for rough, practical estimates.