Glossary
Token (LLM)
The unit an LLM reads and writes - a word or word-piece. Prompts and responses are measured, priced, and rate-limited in tokens.
A token is the atomic unit of text an LLM processes. Tokenization splits text into common word-pieces, so a token is often a whole short word or a fragment of a longer one - roughly four characters of English on average. The model reads its input as a sequence of tokens and generates its output one token at a time.
Tokens matter because they are the meter for everything: providers bill per token, latency grows with token count, and rate limits are set in tokens per minute. Counting and reducing tokens is therefore the lever for controlling the cost and speed of any LLM workload - the basis of token optimization and the token economy.
Govern AI like infrastructure.
Talk to our team about deploying DataStrict across your enterprise stack.