Hi,
I’m curious if pricing of gpt-3.5 doesn’t care about the number of api calls, and is only based on token count. Would these two be priced the same: 2 api calls of 500 tokens in total vs 1 api call of 1000 tokens in total?
You are correct. 500 tokens of the same type as 1000 would be half the cost.
API is billed with a different price for the input and output.
However, because you must prompt the AI to perform its operation, there will be overhead with some near identical tasks - like grammar-checking on same content but split into two.
Just to clarify, is 2 api calls with 500 tokens (both input and output combined) same price as 1 api call with 1000 token combined? Let’s say it is for the same task.
Hiii, @paulrivera22, did you able to find the answer to the question regarding the cost comparison of 1 API call with 1000 tokens vs. 2 API calls with 500 token each.