Skip to main content
Enclave
Glossary

Context Window

The maximum amount of text a single model call can see, measured in tokens. The model knows nothing outside the window — which is why long-term memory systems exist.

Each time an LLM generates, it can only read a bounded input (tens of thousands to a million-plus tokens). The longer the conversation, the more likely earlier content gets pushed out of the window — which is why a bare chatbot starts forgetting things in long chats.

The fix isn't an infinitely bigger window (cost and speed forbid it), but a persistent storage layer outside the window that retrieves relevant content back in on demand — which is exactly what long-term memory and RAG do.

Try it yourself in Enclave

Open it in your browser — no credit card, no install.

Related

More terms

ShareXTelegramLINEWeibo
Context Window · Enclave