# What is a context window?

> A context window is the maximum amount of text an AI model can hold in working memory at once - every message, file, and tool definition in the conversation counts toward it.

Everything you send to a model competes for the same finite space: the system prompt, every message in the thread, uploaded files, tool definitions, and the answer the model is about to write. That space is the context window, measured in tokens rather than words. When a conversation fills it, something has to go, and it is almost always the oldest content.

This is why an assistant can quote your first message perfectly at turn five and have no idea what you said by turn two hundred. The text is still on your screen because the interface stores it, but it is no longer being fed to the model. Context windows vary widely by product and plan, and the number a company advertises is the model's capacity, not necessarily what the consumer chat app actually loads on each turn.

## How this relates to folk

Folk compacts long threads automatically, but the more useful difference is that its memory is not scoped to a thread. What you told it last month is still available next time you text, on whichever surface you happen to be using.

## FAQ

### What happens when you reach the context window limit?

Either the assistant refuses to continue the thread, or it silently drops the oldest messages and keeps answering as if they were never there. The silent version is more common and much harder to notice.

### Does a bigger context window mean better answers?

Not automatically. Attention spreads thinner as context grows, so quality often degrades well before the hard limit. This effect is called context rot, and it is why focused threads usually beat one enormous one.

### Does paying for a subscription increase the context window?

Sometimes, but less than people expect. Paid plans typically raise message rate caps rather than the per-conversation ceiling, so upgrading often does not stop a long thread from filling up.

## Related terms

- [Token Limit](https://www.folk.com/glossary/token-limit), A token limit is the cap on how many tokens - chunks of text roughly three quarters of a word each - an AI model can process in a single request, including both the conversation history and its own reply.
- [Context Rot](https://www.folk.com/glossary/context-rot), Context rot is the decline in AI response quality as a conversation grows longer, caused by the model's attention spreading across more tokens until earlier instructions lose influence.
- [Compaction](https://www.folk.com/glossary/compaction), Compaction is when an AI system summarizes earlier parts of a conversation and replaces them with that summary, freeing space in the context window so the thread can continue.
- [Persistent Memory](https://www.folk.com/glossary/persistent-memory), Persistent memory is an AI's ability to remember information across separate conversations - your preferences, people, and past requests - instead of forgetting everything when a chat ends.

---

Canonical page: https://www.folk.com/glossary/context-window
More about folk (for AI agents): https://www.folk.com/llms.txt · full context: https://www.folk.com/llms-full.txt
folk is a personal AI that lives in your texts (iMessage, Telegram, WhatsApp). Pro $20/mo, Max $100/mo. Made by Nozomio Labs. Sign up: https://www.folk.com/