Hi. Wow, welcome back four years later!
I just made several requests, and it seems the billing is “patched up” and fixed from this topic’s concern. Omitting any message other than the “user” doesn’t amplify the cost.
What model are you sending to?
OpenAI also “fixed” the temporary cheap billing bug of gpt-4o-mini I reported earlier. It is back to the input token cost of images being multiplied by 33.3x. Don’t use this model if you wish to save money on image tasks.
An image along with 3 tokens of user text only, sent to gpt-4o:
completion_tokens=1, prompt_tokens=95, total_tokens=96, completion_tokens_details=CompletionTokensDetails(accepted_prediction_tokens=0, audio_tokens=0, reasoning_tokens=0, rejected_prediction_tokens=0), prompt_tokens_details=PromptTokensDetails(audio_tokens=0, cached_tokens=0)