In short
The unit a model reads and writes in. Roughly three quarters of a word. Models are priced and limited by tokens, not pages.
Before a model reads text it breaks it into tokens: pieces of words, whole short words, punctuation. A thousand tokens is about seven hundred fifty English words. Everything about a model is measured this way: how much it can see at once, how much a call costs, how long an answer can be.
You rarely need to count them. You need to know they exist, because they explain two things that surprise people: why a long document costs more than a short one, and why an agent loads only what it needs for the run instead of everything it has ever seen.
Example
What it looks like.
token.mdplain text
A two-page brief is about 1,000 tokens. An instruction file about 800. A long meeting transcript about 12,000. All three fit in one run with room to spare on a current model.Related
Read next.
- context windowEverything the model can see on one run: the instructions, the files it was given, the conversation so far. Outside the window, it does not exist.Read the definition
- modelThe intelligence an agent calls when it needs to read, write, or decide. Replaceable. When a better one ships, the same instructions run on it and the results improve.Read the definition
- inferenceRunning a model to get an answer. It costs money per call in the cloud, or electricity on your own machine. Where it runs is a choice.Read the definition