document OCR / layout analysis / table recognition
Surya is a deep-learning OCR model that performs text recognition, layout analysis, reading order, and table recognition in 90+ languages, packaged as a Python library.
Same problem, same approach. Swapping one for another is a config change, not a rewrite.
drop-in peer for lightweight PDF-to-Markdown with strong table parsing.
OCRFlux is a lightweight yet powerful multimodal toolkit that significantly advances PDF-to-Markdown conversion, excelling in complex layout handling, complicated table parsing and cross-page content merging.
drop-in peer from Zhipu AI with similar OCR/layout/table coverage.
GLM-OCR: Accurate × Fast × Comprehensive
drop-in peer with mature, widely-adopted OCR/layout/table toolkit and extensive language support.
Turn any PDF or image document into structured data for your AI. A powerful, lightweight OCR toolkit that bridges the gap between images/PDFs and LLMs. Supports 100+ languages.
Solves the same problem with a different architecture or at a different layer. Expect to rewrite the integration.
modular framework to combine multiple document AI models rather than a single integrated one.
A Repo For Document AI
choose if you need math formula recognition or Markdown output with small CPU-friendly models.
An Open-Source Python3 tool with SMALL models for recognizing layouts, tables, math formulas (LaTeX), and text in images, converting them into Markdown format. A free alternative to Mathpix, empowering seamless conversion of visual content into text-based representations. 80+ languages are supported.
high-level OCR library with multiple model backends, covering OCR-only tasks.
docTR (Document Text Recognition) - a seamless, high-performing & accessible library for OCR-related tasks powered by Deep Learning. Ongoing development and maintenance by t2k.
easy-to-integrate OCR library without layout/table analysis.
Ready-to-use OCR with 80+ supported languages and all popular writing scripts including Latin, Chinese, Arabic, Devanagari, Cyrillic and etc.
uses external vision-LLM APIs for flexible OCR/extraction rather than a local model.
OCR & Document Extraction using vision models
research-grade anchor-prompting parser, less mature but distinct architecture.
The official repo for “Dolphin: Document Image Parsing via Heterogeneous Anchor Prompting”, ACL, 2025.
rich research toolbox with many text detection/recognition models for customization.
OpenMMLab Text Detection, Recognition and Understanding Toolbox
fast LLM-based document parser for structured extraction instead of low-level OCR.
A fast, helpful, and open-source document parser
ONNX-based OCR toolkit with fast CPU inference and multiple language bindings.
📄 Awesome OCR multiple programing languages toolkits based on ONNX Runtime, OpenVINO, MNN, PaddlePaddle, TensorRT and PyTorch.
An earlier or more limited way to do the job. Still the right call when you need something small, proven, or CPU-only.
narrower toolkit for layout detection only, without OCR/reading order.
A Unified Toolkit for Deep Learning Based Document Image Analysis
exclusively table extraction, no OCR or layout analysis.
Table Transformer (TATR) is a deep learning model for extracting tables from unstructured documents (PDFs and images). This is also the official repository for the PubTables-1M dataset and GriTS evaluation metric.
JavaScript OCR engine for browser/Node.js, simpler layout capabilities.
Pure Javascript OCR for more than 100 Languages 📖🎉🖥
native Swift OCR for Apple platforms, no layout/table support.
Fast and simple OCR library written in Swift
legacy Python OCR pipeline for historical documents, no modern table/layout.
Python-based tools for document analysis and OCR
Go binding for Tesseract, simple OCR without layout analysis.
Go package for OCR (Optical Character Recognition), by using Tesseract C++ library
recognition-only model for cropped text images, no detection or layout.
Text recognition (optical character recognition) with deep learning methods, ICCV 2019
battle-tested legacy OCR engine, lighter weight but no deep-learning table/layout.
Tesseract Open Source OCR Engine (main repository)
Does the same job, but ships as an app. Useful to a person, not swappable into a codebase.
end-user cross-platform GUI tool, not an embeddable library.
树洞 OCR 文字识别（一款跨平台的 OCR 小工具）
experimental Android app, not a library for integration.
Experimental optical character recognition app
These projects were analysed and named surya among their alternatives. The relationship is not symmetric — how surya rates them is a separate judgement, made when surya is analysed in its own right.
calls surya “Python OCR/layout-analysis engine with reading order and table recognition; choose when you need layout-aware extraction beyond plain OCR.”
Tesseract Open Source OCR Engine (main repository)
calls surya “Choose when you need OCR plus layout analysis and reading order in a single, compact library that works across 90+ languages.”
Ready-to-use OCR with 80+ supported languages and all popular writing scripts including Latin, Chinese, Arabic, Devanagari, Cyrillic and etc.
calls surya “Choose for broader language support (90+) and an actively maintained OCR pipeline with full layout analysis.”
GLM-OCR: Accurate × Fast × Comprehensive
calls surya “choose when you need a library with OCR plus layout, reading order, and table recognition in 90+ languages.”
Multilingual Document Layout Parsing in a Single Vision-Language Model
calls surya “Python OCR library with layout analysis and 90+ languages, choose for modern multilingual Python OCR.”
树洞 OCR 文字识别（一款跨平台的 OCR 小工具）
calls surya “if you need multilingual OCR and layout analysis for images/PDFs, not office format conversion”
Get your documents ready for gen AI