A fast, helpful, and open-source document parser for efficient local processing. LiteParse excels at extracting text with precise bounding box information from various document types, including PDFs, Office documents, and images. It offers flexible OCR capabilities, supporting a bundled Tesseract engine and easy integration with external HTTP OCR servers.

Overview

LiteParse is a standalone, performance-focused document parsing tool designed for speed and efficiency. It provides high-quality spatial text parsing with bounding boxes directly on your machine, eliminating the need for proprietary LLM features or cloud dependencies. This makes it an ideal choice for applications requiring rapid local document analysis.

Key Features

• Fast Spatial Text Parsing: Utilizes PDFium for rapid text extraction and accurate spatial positioning.
• Flexible OCR System: Includes a zero-setup, bundled Tesseract OCR engine, with seamless integration for external OCR servers (e.g., EasyOCR, PaddleOCR) via a defined HTTP API.
• Screenshot Generation: Produces high-quality page screenshots, crucial for LLM agents and visual document analysis.
• Comprehensive Input Support: Handles PDF, DOCX, XLSX, PPTX, and various image formats by leveraging LibreOffice and ImageMagick for automatic conversion to PDF.
• Precise Bounding Boxes: Outputs structured JSON containing text along with its precise bounding box coordinates for detailed analysis.
• Multi-Language & Multi-Platform: Accessible from Rust, Node.js/TypeScript, Python, and the browser (WASM), supporting Linux, macOS, and Windows.
• Multiple Output Formats: Generates output in both structured JSON and layout-preserved plain text.

Typical Use Cases

• Local Document Analysis: Quickly extract text and layout information from documents without relying on cloud services.
• Data Extraction from Scanned Documents: Leverage the integrated or external OCR capabilities to process scanned PDFs and images.
• Building Document Processing Pipelines: Integrate LiteParse into existing workflows for automated document ingestion and analysis.
• LLM Agent Augmentation: Provide LLM agents with structured text data and visual context (screenshots) for more informed decision-making.
• Batch Document Processing: Efficiently parse large collections of documents using the provided batch processing tools.
• Development of Cross-Platform Document Tools: Utilize its multi-language bindings to build applications for various environments.