What is a context window?
A context window is the maximum amount of information, measured in tokens, that a language model can consider at one time, including instructions, conversation history, retrieved documents and its own output. Anything beyond the window is not seen by the model.
Why it matters
| Issue | Effect |
|---|---|
| Too little context | Model misses facts |
| Too much context | Higher cost, slower, may miss details |
Managing context
Retrieve only relevant information, summarize long threads and store long-term facts in memory or a knowledge graph.
Larger windows
Windows have grown from thousands to millions of tokens, but relevance still matters more than size.
