The OpenAI pricing page states that “In the simplest case, if your prompt contains 10 tokens and you request a single 90 token completion from the davinci engine, your request will use 100 tokens and will cost $0.002.”
This seems to imply that any api request carries a 1K token charge minium. Even if prompt+completion is 100 tokens, you will be charged for 1K tokens. By the same logic 1001 tokens would cost as much as 2000 tokens.
Can someone confirm there’s always a minimum 1K token charge?