Token
Definition: A token is the basic unit an LLM processes: a chunk of a word (often ~4 characters in English). Text is split into tokens before processing.
Models bill and limit usage per token (input + output). The longer the text, the more tokens it consumes.
Frequently asked questions
How many tokens are in a word?
In English, ~1 token ≈ 4 characters, about 0.75 words. Rule of thumb: 1,000 words ≈ 1,300–1,600 tokens.
Why is usage billed per token?
A model's compute cost is proportional to the number of tokens processed (input + output). APIs therefore bill per token, with separate input and output rates.