paper2md: PDF-to-Clean-Text Extraction Service

Spitale G. (2026) borant.eu [software]

Abstract

Turns a scientific PDF into clean, reflowed text using a full document-layout model — reading order reconstruction, stripped tables/headers/footnotes, references and back matter removable on request. Public no-login web page, or an API with issued keys for programmatic access (e.g. feeding corpora into AutoCode).

Who it is for: Anyone holding a PDF. No account, no sign-up, nothing to accept.

Source code: https://github.com/that-ugly-cat/paper2md

Cite

Spitale, G. (2026). paper2md: PDF-to-Clean-Text Extraction Service. borant.eu.

@misc{spitale2026papermd,
  author  = {Spitale, G.},
  title   = {{paper2md: PDF-to-Clean-Text Extraction Service}},
  publisher = {borant.eu},
  year    = {2026}
}

Published in

Year

2026

DOI

Topics

borant, research, software