#Can someone pls explain to me why this
3 messages · Page 1 of 1 (latest)
GPT hisself is speculating that this is the reason: ```The "18" prompt tokens are the number of tokens in the input message. In this case, the input message consists of two parts: the system message and the user message. Here's the token breakdown:
System message: "You're a helpful assistant." (6 tokens)
User message: 'hello' (1 token)
Additional tokens: role, content, model, messages, stream, temperature, and punctuation (11 tokens)
Total tokens: 6 (system) + 1 (user) + 11 (additional) = 18 tokens.```. Legit or hallucination?
Seems legit, it has to send all relevant prompt information with each query