I’ve set up a sample API server and integrated it with ChatGPT actions. Everything is functioning correctly with all the endpoints, except for one. This particular endpoint responds with a PDF, but ChatGPT is receiving an empty object. In an attempt to resolve this, I modified the endpoint’s response format to provide a download link for the PDF. However, ChatGPT still cannot process these files. It’s worth noting that I can access the PDFs from Postman and Swagger, which suggests that the API itself is working as expected. I’m wondering if there are specific metadata settings or headers that need to be configured for ChatGPT to handle PDF responses correctly from the API.