On the OpenAI Playground (https://platform.openai.com/playground/assistants 3), set the model to GPT-4o. Then, use the following link to upload the image: https://aidoctor.world/upload/2024-06-13/o6K026lswlFF1EWcZe2eWI492-XM-IMG_6636.png.
Please note that GPT-4o may not be able to interpret the content of the image.
However, it performs well with the same image on https://platform.openai.com/playground/chat?models=gpt-4o 1.
Why?
How to solve this problem?