Яндекс.Метрика

LLM token

LLM token explained

The unit of text a language model counts in: a word or part of one. Tokens measure the context window, the length of a request and what an LLM costs to run

An LLM token is a chunk of text a language model works with: a whole word, part of a word, a punctuation mark or a space. Models read and write text in tokens, so tokens are how you measure:

  • The size of the context window
  • The length of a prompt and a response
  • What a model provider charges per request
  • Usage limits

Russian text usually breaks into more tokens than English text with the same meaning.

Don't confuse it with a Pachca API token, the secret key an agent or integration uses to authorize API requests. Pachca's tools help agents save LLM tokens: Pachca CLI returns command details only when asked, and the docs are available in Markdown without extra markup.

Related terms