Chatgpt error: No text could be extracted from this file

I am trying to upload a handwritten document scanned into a pdf.
Its perfectly legible and scannable on other ocr software easily.
however iin chatGPT I get this error.
no matter what model I use. Also I have plus account.

“No text could be extracted from this file.”

anyone else getting this?

That behavior is expected. ChatGPT only extracts text from PDFs that already contain searchable text. The PDF you uploaded is a scanned handwritten document. Even though it is clearly readable and other OCR software can recognize it, ChatGPT itself does not perform OCR directly on uploaded PDF files. Instead, the PDF must first be processed by the same OCR software you describe, or other that can enhance the PDF.

That OCR software (or OCR that accompanies the scanning solution) adds searchable text to the PDF. Once the PDF has searchable text, ChatGPT can then extract and use it through OpenAI’s file search and vector store system. This limitation applies to all ChatGPT models, even if you have a ChatGPT Plus account, despite the user interface appearance of “attachment”.

However, there is a new feature available through OpenAI’s API and the Responses endpoint. This feature allows you to attach a PDF file directly as part of a message. When you attach a PDF this way, the system automatically extracts text and also visually processes each page. This rendering and placement for GPT-4o vision helps the AI better understand the document. As a result, ChatGPT can handle tasks that were previously difficult with a chunking-search technique, such as summarizing an entire PDF document, even if it was not originally searchable.

Getting the same issue as well…

Hi, welcome to the community!

  • Right click on PDF file + Open with + select your browser

  • Click on Printer icon
    polepole-1

  • Select Microsoft Print to PDF
    polepole-2

  • Click on Print
    polepole-3

  • Save with a new file name.

  • Upload new file in ChatGPT

This seems to be an artificial limitation. If I convert the pdf to png, ChatGPT will extract the text perfectly.