My API prompt token is exceeding to some unbelievable numbers. some user is using some kind of injections in order to alter the max number of tokens and model. while i am chatgpt 3.5 turbo and my max token limit is 150…

Welcome to the forum.

The token usage numbers look fairly low, it looks like an application error, can you post a snippet of the code that is making the API calls?

1 Like

You should also cancel your current API keys if you suspect something untoward is coccuring.

1 Like

Here’s a good test: do you even have access to GPT-4 yourself?

May be similar billing database problems as this: