Surya — The Complete Guide: OCR, Tables and LaTeX Equations from Scanned Documents
Surya, from Datalab, is a Python document recognition toolkit that excels precisely at the hard parts of an engineering page: tables and equations. It runs on NVIDIA GPUs, on CPU or on Apple Silicon.
Surya
Datalab
Python OCR toolkit for 90+ languages with layout and reading-order analysis, tables and equations as LaTeX; models free for personal use and research.
Get it from the vendorWhat you get
- Multilingual OCR — more than 90 languages.
- Layout analysis — including the correct reading order on multi-column pages.
- Table recognition — rows, columns and merged cells.
- Equations as LaTeX — inline in the text, ready to copy.
- Flexible hardware — NVIDIA, CPU or Apple Silicon.
What is free and what is not
The code is under Apache 2.0, but the model weights — without which the tool does not work — use a modified Open Rail-M licence. They are free for research, personal use and companies with under 5 million US dollars in funding or revenue. A larger company using the tool commercially needs a paid licence from Datalab. There is no regular graphical interface, and installation is aimed at Python users.
Related tools
For an alternative fully under Apache 2.0 — PaddleOCR. For converting documents to Markdown under MIT — Docling. For capturing a single equation from a screenshot — Mathpix Snip.
PaddleOCR
PaddlePaddle (Baidu)
Open-source Python OCR and document parsing toolkit: text recognition in dozens of languages, page layout, tables and PDF to Markdown or JSON.
Get it from the vendor
Docling
Docling Project (originally IBM Research)
Open-source tool that converts PDF, DOCX and scanned images to Markdown or JSON, with table structure recognition and OCR, running locally.
Get it from the vendor
Mathpix Snip
Mathpix
Captures equations, tables and text from a screenshot and converts them to LaTeX, Markdown or CSV; the free plan gives 10 images and 10 PDF pages a month.
Get it from the vendor
docTR
Mindee
Open-source deep-learning OCR library: text detection and recognition in PDFs and images, including rotated pages and table region detection.
Get it from the vendorThe bottom line
For a student, a researcher or a small office, one of the best tools for turning scanned standards and textbooks into text with tables and equations. At a larger company, check the licence before use.