Selection guide / editorial
Turn PDFs into usable documents
The target can be searchable text, reading-order Markdown, tables or a structured document. Choose and evaluate that contract explicitly.
Check the PDF
Identify born-digital pages, scans and mixed files. Preserve the source document and page numbers so missing text is easy to trace.
Pick the output
Use Markdown for prose workflows. Keep structured table cells and page regions when downstream software needs reliable table relationships or coordinates.
Inspect difficult pages
Check multiple columns, footnotes, rotated pages, formulas and table headers. Average text accuracy can hide broken structure.
Evaluate the pipeline
OmniDocBench, olmOCR-bench and ParseBench are catalogue candidates for different document-parsing aspects. Our ParseBench adapter is still pending; source-reported scores are shown separately.
A minimal document conversion
Illustrative Docling integration, following its upstream quickstart; this snippet is not a benchmark run.
from docling.document_converter import DocumentConverter
result = DocumentConverter().convert("sample.pdf")
markdown = result.document.export_to_markdown()
print(markdown)