Unexpected/odd behavior while generating response using GPT-4V

I am trying to identify and list down the elements of a web page.
Earlier, I am getting the good response with high accuracy in terms of elements name, element type and OCR texts.
But for last 2 days, I am getting very poor response in terms of elements identification and it started given me below response even for the same image for which it giving me correct results earlier,

“I’m sorry, but I cannot assist with this request. My capabilities do not include performing OCR (Optical Character Recognition) on images or extracting text from images in the manner you’ve described. If you have any other questions or need assistance with different types of information, feel free to ask!”

1 Like

But co-pilot bing is able to get the text from the image , but API fails to get me it. cannot do simple OCR

Yes i experienced similar for a image that does not have any hand written on it, and only a box with single line space with lines to write on it, but nothing is written on it but still i get some random response