AI Intensify: Practical AI for small business owners, explained plainly, with the real costs.

© 2026 AI Intensify

How models work

What is a token in AI?

Also called

  • tokens
Last checked How we check this

Definition

A small chunk of text, often part of a word, that language models read and write, and that providers use to measure usage and price.

Key takeaways

  1. A token is a small chunk of text, often part of a word, that language models read and write.
  2. How many characters fit in a token depends on the model and the language.
  3. Token counts affect what you pay on usage-based plans and how much fits in the context window.

In plain English

Models do not see words the way people do: they break text into tokens. How many characters fit in a token depends on the model and the language, so the same text can count differently on different models. Token counts affect two things you will notice: how much you pay on usage-based plans and how much fits in the context window. Example: a long contract pasted into a chat uses far more tokens than a short question.

A small business example

Picture a bookkeeper on a usage-based plan who pastes whole contracts into a chat for every small question. Each long paste uses far more tokens than a short question, so the bill grows quietly. Pasting only the clause that matters keeps costs down and leaves more room in the context window.

Questions

Why do tokens matter on a flat monthly plan?

Even on a flat plan, tokens decide how much fits in the context window, the amount of text a model can consider at once. On usage-based plans they also decide what you pay.

How we check this

Each definition is written in plain English from named sources, such as the EU AI Act, NIST and OECD publications, and the providers' own documentation. Those sources are listed on each term page, with the date the term was last checked. Definitions are kept short on purpose, so the sources are the place for the full detail. Meanings shift as tools change, so a definition that looks out of date is worth flagging through the Tell us about a change link on any term page.