Description
Copying formulas and tables from paper PDFs produces garbage, and regular OCR can't read math. Nougat is Meta's neural OCR model for academic documents, turning paper PDFs into Markdown with LaTeX math.
It handles formulas, tables and multi-column layouts well, converting in batches from the command line.
Math:LaTeX output.
Tables:Structure kept.
Markdown:Easy editing.
Batch:CLI.
It handles formulas, tables and multi-column layouts well, converting in batches from the command line.
Features
Math:LaTeX output.
Tables:Structure kept.
Markdown:Easy editing.
Batch:CLI.
