Currently, if the OpenAI API returns a 400 bad request error, then the backend will log it and return a 500 server error. We should investigate to see if we can handle such errors in a better fashion and potentially return a 400 error to the user if the reason for the bad request is because of user input.
Currently, if the OpenAI API returns a 400 bad request error, then the backend will log it and return a 500 server error. We should investigate to see if we can handle such errors in a better fashion and potentially return a 400 error to the user if the reason for the bad request is because of user input.