There's also a "how much it reads to produce that data."
If you go to https://beta.openai.com/account/usage
you will see things like:
text-davinci, 5 requests
187 prompt + 218 completion = 405 tokens
Note at the bottom of
https://beta.openai.com/playground/p/default-grammar?model=t... - you will see a counter in the lower right corner. The default that this playground model opens with shows 15 tokens in the prompt and the friend chat (
https://beta.openai.com/playground/p/default-friend-chat?mod... ) starts out with 28 tokens. Submitting it, and the conversation now has 41 tokens in it... and adding another line, 50 tokens... and the response, now 97 tokens. And so on.