How it works
Context is measured in tokens, roughly three-quarters of a word each. When a conversation grows past the window, older parts drop out or get summarized, which is why long chats can forget early details.
Why it matters
Larger context windows let a model work with long documents and long conversations. Each model in UncutGPT's picker has its own limit.
