Confused about the costs per request

Hello there. I´m working on a project and honestly I´m really confused about the costs per request from my gpt4o api.
I want to analyse an image which has aprox. 681.944 tokens.
My text prompts are about 579 tokens.

There are several calculator for those requests, but I really can´t get a clear result for me.
Can someone pls help me how to determine how much I have to pay for each request?

I tried to get the costs over the usage monitor from openai. For one request there are 640 context tokens and 476 geberated tokens, summary 1116 tokens.
The costs are $0.01 …so 1 Cent for each request? Is that right?


I arrive at the same price, i.e. 1 cent per request.

I assume you may have already seen this guidance for vision cost calculation? If not, then it’s a useful reference in case of doubt in addition to the vision pricing calculator on OpenAI’s website.

Any more specific questions, just let us know.

1 Like