← Latest products

OCR pipeline for ML training (tables

diagrams, math, multilingual). Multi-modal OCR pipeline optimized for ML training (text, figure, math, tables, diagrams). It’s designed to process complex academic materials — including math formulas, tables, figures, and multilingual text — and output clean, structured formats like JSON and Markdown.

Open source

674 stars · GitHubexternal site

Visit github.comexternal site

Reactions (1)

Similar products

by description similarity

  1. A highly extensible framework for building OCR systems

    github.com

    A highly extensible and customizable framework for building OCR systems.

    Hacker News16 pts

  2. PDF to MD by LLMs

    github.com

    Extract Text/Tables/Image Descriptives by GPT4o

    Hacker News191 pts

  3. Open-Source Document Extraction Tool

    github.com

    Open-source spreadsheets platform for deep research and document processing

    Hacker News13 pts

  4. Surya

    github.com

    OCR and line detection in 93 languages

    Hacker News11 pts

  5. Autofit2

    github.com

    End-to-end pipeline for multilingual text classification

    Hacker News28 pts