Respond in JSON format to divide the different captions in frontendWhen OpenAI includes the possibility to use the respond format this will be implemented
Providing 4 different captions at onceNot only giving one caption but four. The challenge is to get it back as a JSON to prepare it for the frontend.