Burkov builds near-perfect OCR pipeline turning papers and textbooks into Markdown

burkov · x · 2026-09-10

Andriy Burkov says his ChapterPal OCR pipeline is virtually perfect: it handles scientific articles and textbooks, outputting Markdown with illustrations, complex tables (down to the original LaTeX), and LaTeX equations. As a solo developer with limited validation resources, he invites users to stress-test challenging PDFs via the Premium OCR option, with a 7-day free trial available.

Original post →

More from Apps

Apps channel →