---
title: "Token and context window – AI limits explained | Sweet Lava"
description: "A token is the smallest piece of text a model processes; the context window caps how many fit in one call. See why the limit shapes AI tools."
url: https://sweetlava.eu/en/glossary/token-and-context-window/
lang: en
alternate_pl: https://sweetlava.eu/pl/slownik/token-i-okno-kontekstowe/
---

SEO & AI 

# What is Token and context window?

A token is the smallest piece of text a language model processes: a word fragment, a short whole word or a character. The context window is the limit of tokens a model handles in a single call; it has to fit the instruction, any attached material and the generated answer.

Models don't read letters or words – they read tokens. In English a token averages a few characters; words in other languages often split into more tokens, so the same text can "cost" more outside English. This translates directly into money: model providers bill usage in tokens, separately for input and output.

The context window sets how much the model sees at once. When a conversation or document exceeds the limit, the oldest part drops out: hence assistants that lose earlier arrangements in a long conversation. It's also why sensible implementations don't feed the model the entire company knowledge base: the RAG technique passes it only the fragments that match the question, which is cheaper and more effective.

**For buyers:** when pricing an AI implementation, ask how the vendor estimates token costs at your real scale (number of queries, document length) and what happens with conversations longer than the context window. These two questions quickly separate an experienced team from an enthusiast.

## Related services

- [ AI integration services → ](https://sweetlava.eu/en/services/ai-integration-services/)

## Related articles

AI · 12 June 2026

## [ Implementing AI in a company, step by step – what our own projects taught us ](https://sweetlava.eu/en/blog/ai-implementation-step-by-step/) 

How AI implementation works in practice – from a data audit, through a pilot with a measurable goal, to team enablement. With concrete price ranges and a list of situations where AI isn't worth it.

## Related terms

- [LLM (large language model)](https://sweetlava.eu/en/glossary/llm/)
- [Prompt & prompt engineering](https://sweetlava.eu/en/glossary/prompt-engineering/)
- [RAG (retrieval-augmented generation)](https://sweetlava.eu/en/glossary/rag/)

## Wondering what AI could change in your company?

We implement AI where it makes measurable sense and teach your team to use the tools. We'll also tell you straight when it's not worth it.

[ Get in touch ](https://sweetlava.eu/en/contact/?source=glossary-token-context-window)