IBM launches Docling, a free Python library to convert any document to data

mdancho84 · x · 2026-08-21

IBM has released Docling, a free Python library designed to parse various document formats, especially PDFs, into structured data. It features PDF backends, layout analysis models, vision-based table formatters, and OCR capabilities. It can convert complex documents like arXiv papers directly into Markdown format for seamless integration with the Gen AI ecosystem.

Original post →

More from Infra

Infra channel →