Hi Folks,
We are using the OpenAI API in our project for various use cases such as chatbots and automation tasks, as well as for updating real-time data.
However, I am encountering an issue where the responses from the OpenAI API (using the GPT-4o model) are irrelevant and inconsistent compared to the valid responses I receive from the ChatGPT app using the same model and prompt.
I am unsure of the reason for this discrepancy despite using the same GPT-4o model and prompt in both instances.
Has anyone else experienced similar issues? Are there any known differences in how the API processes requests compared to the ChatGPT app?
Additionally, I would appreciate any insights or suggestions on how to troubleshoot and resolve this issue. For example, are there specific API settings or parameters that might affect the response quality?
Thank you in advance for your assistance!