I needed tools for better PDF parsing (the ability to parse tables properly) for an RAG application.

Currently, I am using LlamaParse for parsing the pdf document to markdown but it's not a self-hosted service, and I need something self-hosted as I am working with confidential data. I also tried Llmsherpa but the results were nowhere close to LlamaParse at various points like extracting some complex tabular structures.

Are there any other open-source tools I can use for PDF parsing that can handle tables and sections easily? The documents that I need to parse have some tables that are complex to read, and the RAG application performance relies on context extracted by PDF.

I also came across unstructured.io for PDF parsing but have not tried it yet will be checking it next. One thing I would like to ask before investing time in setting up an environment locally. In Official Documentation of Unstructured.io setting up unstructured.io locally on Windows will add additional challenges, so has anyone tried it locally and is it completely open if I set its environment locally will it mean all functionalities will be performed on my local machine or not?

Are there other ways to solve this problem that I might have missed?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.