# What is a 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.

Models do not read words, they read tokens. A token is roughly three quarters of an English word, so a thousand words costs somewhere near thirteen hundred tokens. Punctuation, code, and non-English text tokenize less efficiently, which is why a page of code eats far more budget than a page of prose.

The limit applies to the whole request, not to your latest message. That is the detail almost everyone gets wrong. A three-word follow-up still gets appended to the entire preceding conversation before it is sent, so a tiny message can be the one that tips a long thread over the edge. Tool definitions and connected integrations are charged against the same budget, which is why disabling unused tools can buy back real room.

## How this relates to folk

Folk watches its own token budget and compacts before it runs out, so you are not the one tracking it. There is no counter to babysit and no point at which the conversation refuses to continue.

## FAQ

### How many words is a token?

Roughly three quarters of a word in English, so 1,000 tokens is about 750 words. Code, punctuation, and other languages tokenize less efficiently and consume more of the budget per character.

### Why does it say maximum length when my message was short?

Because the limit counts the entire conversation, not your last message. Every previous exchange is resent with each turn, so a short message can be the one that pushes the total over the ceiling.

### Do uploaded files count toward the token limit?

Yes. Files, images, tool definitions, and system instructions all consume the same budget as your messages, which is why a thread with several attachments fills up much faster than a plain text one.

## Related terms

- [Context Window](https://www.folk.com/glossary/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.
- [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.
- [Large Language Model (LLM)](https://www.folk.com/glossary/large-language-model), A large language model (LLM) is an AI trained on vast amounts of text to understand and generate human-like language, powering chatbots, assistants, and AI agents.

---

Canonical page: https://www.folk.com/glossary/token-limit
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/