hire us
Glossary·
Minsk · WarsawHire us
Context window
The maximum amount of text a model can hold in view at once — the prompt, the retrieved documents, the conversation so far, and the answer it is generating.
Why it matters
It sets a hard ceiling on what the model can consider, and it is what you pay for. Long conversations and large documents hit the limit, and the earliest content is what gets dropped.
How we approach it
We budget the window deliberately: retrieve the few relevant passages instead of stuffing whole documents, summarise long histories, and cache the stable part of a prompt to cut cost on every call.
