How text is counted
A token can be a whole short word, part of a longer word, or punctuation. Even spaces affect how text is split, so token counts are not the same as word counts.
Limits and costs
A model’s context window is the maximum number of tokens it can handle in one request. Your prompt, earlier messages, included documents, and the reply all use that space. The limit varies by model and service.
AI services often charge separately for what you send (input tokens) and what the model produces (output tokens). More text can increase cost and response time, and make important details harder for the model to find.
These text tokens are different from access tokens, which grant permission to use a service.