How to close the API cache pricing?

API
1

When I used gpt-4o api, the cache service run automatically. I only need to do that task one time. But the cache service charge me a lot of money. And I do not know how to close this service!!! This made me very angry!!!

The image shows a breakdown of expenses on February 5th, totaling $9.92, for various GPT-4 usage, with specific costs for inputs, outputs, and cached inputs. (Captioned by AI)
The image shows a breakdown of expenses on February 5th, totaling $9.92, for various GPT-4 usage, with specific costs for inputs, outputs, and cached inputs. (Captioned by AI)416Ă—208 10.4 KB

2

https://platform.openai.com/docs/guides/prompt-caching#frequently-asked-questions

  1. **Is there a way to manually clear the cache?**Manual cache clearing is not currently available. Prompts that have not been encountered recently are automatically cleared from the cache. Typical cache evictions occur after 5-10 minutes of inactivity, though sometimes lasting up to a maximum of one hour during off-peak periods.
3

You are misinterpreting the usage report.

Cached input is saving you money. It is a 50% discount on the parts of message inputs that you have recently sent before.

That $8 could have been $16 of just “input”.

You can manage the total expenditure by looking at the length of previous conversation you continue to re-send when continuing a session. If you’d save more by a large minimizing rather than having a cache discount for the “old chat” part, you can delete the oldest parts of conversational memory.

2 Likes