← Tutorials17 / 22

What is an LLM?

What is the context window?Tap to reveal

The maximum number of tokens the model can consider at once: system prompt, conversation, documents and its own output combined.

Longer contexts cost more: naive attention compares every token with every other (quadratic), and the KV cache of past tokens grows with length. Anything beyond the window is simply not seen.

← PreviousNext card →