There’s a big difference between creating Custom GPTs, uploading files, and asking questions, and using the
API to embed the file and then ask the question.
Is the embedding engine used by ChatGPT, or GPTs, different from the embedding engine provided by the API?
Is there any way to use the same one that GPTs uses, since GPTs seems to perform better?
For reference, I tested using both the text-embedding-3-small model and the ada_2 model.
Thanks in advance for any answers.