Incorrect tabular data extraction for handwritten docs when submitted as PDF while correctly extracting accurate data when submitted as an image. #320

Open
opened 2026-02-16 00:17:28 -05:00 by yindo · 1 comment
Owner

Originally created by @UsamaAmjad03 on GitHub (Nov 5, 2024).

Describe the bug
I am submitting a File containing the tabular data with handwritten text in it with model openai gpt-4o, but when being submitted as an image (jpeg, png etc) it gives accurate result but when submitted as a PDF file it give wrong values inside table.

Job ID
4ab24faf-4ef1-4cba-bc04-96f8f1c6de09 (For PDF that gave wrong result)

f4fb4c09-32c3-4a90-bfee-d80e32e81c5b (For image that gave accurate values)

Client:
Please remove untested options:

  • Python Library
  • API
  • Frontend (cloud.llamaindex.ai)
Originally created by @UsamaAmjad03 on GitHub (Nov 5, 2024). **Describe the bug** I am submitting a File containing the tabular data with handwritten text in it with model openai gpt-4o, but when being submitted as an image (jpeg, png etc) it gives accurate result but when submitted as a PDF file it give wrong values inside table. **Job ID** 4ab24faf-4ef1-4cba-bc04-96f8f1c6de09 (For PDF that gave wrong result) f4fb4c09-32c3-4a90-bfee-d80e32e81c5b (For image that gave accurate values) **Client:** Please remove untested options: - Python Library - API - Frontend (cloud.llamaindex.ai)
yindo added the bug label 2026-02-16 00:17:28 -05:00
Author
Owner

@BinaryBrain commented on GitHub (Nov 8, 2024):

We don't support handwritten text if that's the problem.

@BinaryBrain commented on GitHub (Nov 8, 2024): We don't support handwritten text if that's the problem.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: run-llama/llama_cloud_services#320