Global

This package contains an OCR engine - libtesseract and a command line program - tesseract.

LiteParse is a standalone OSS PDF parsing tool focused exclusively on fast and light parsing. It provides high-quality spatial text parsing with bounding boxes, without proprietary LLM features or cloud dependencies. Everything runs locally on your machine.

Fast Rust library for PDF inspection, classification, and text extraction. Intelligently detects scanned vs text-based PDFs to enable smart routing decisions.

纯 C# PP-OCRv6 推理库：多平台手写 Kernel / GEMM 等算子、超高性能、低内存需求、超高准确率。 自带托管 ONNX 解释器，不依赖 Paddle Inference、ONNX Runtime 或 OpenCV 原生库。

The maintainFormat option tries to return the markdown in a consistent format by passing the output of a prior page in as additional context for the next page. This requires the requests to run synchronously, so it's a lot slower.

pdf2htmlEX is no longer under active development. New maintainers are wanted.