# Let's say I have 100 prompt tokens that will result in 50 completion tokens, that is 150 total tokens. By sending my 100 tokens to the chat completion API did I just use 150 tokens from tokens-per-minute limit or did I just use 100?