Im looking to moderate images prior to sending them to the gpt-4o or vision endpoint. I had a warning last week from OpenAI and I dont want to get banned. If there is not a way to moderate images, what approach should I consider? Any ideas?
Thank you in advance
Does this function with image input? I think it’s solely applicable to text.
I’m sorry. Yeah. The endpoint is only for text. I clearly didn’t read this properly.
Not sure how to moderate vision with OpenAI.
May just need to use a third-party model
The same way they detected a violation for the warning in question, just as with the text, as for now it seems like just making an existing internal API or it’s version available publically