For GPT4o function calling is faster by a factor of 4-5x compared to structured formatting - even though I can effectively force a function call to provide the same output.
is there anyway to speed the process up at all, and is the structured formatting substantially more expensive given that it is slower, I assume it uses more compute
Also in general, for python is it recommended to utilize the SDK/openai library or just rely on requests? (currently I just use requests, but wanted to understand if there is a strong reason to switch to library)