marker

Convert PDF to markdown + JSON quickly with high accuracy

41.0k
Stars
Python
Language

💡 Why It Matters

This project addresses the need for accurate document conversion to markdown and JSON, handling a variety of formats including PDF and DOCX. Unlike alternatives like MinerU or docling, it offers a hybrid approach that optimises performance on both CPU and GPU, ensuring high throughput without sacrificing accuracy. Backend engineers and ML/AI teams will find it particularly beneficial, given its extensibility and ability to integrate LLMs for improved results. However, with nearly 41,000 stars and a last commit in late 2026, it lacks a meaningful growth history, which may raise concerns about its long-term viability. It may not be suitable for teams requiring a well-established tool with extensive community support.

🎯 When to Use

This is a strong choice for projects that require fast and accurate document parsing, especially when working with diverse file types. Consider alternatives like PaddleOCR or LightOnOCR for simpler OCR tasks or when extensive community support is essential.

👥 Team Fit & Use Cases

It often integrates into systems for converting legacy documents into searchable formats, such as transforming old reports into markdown for internal documentation or building features that require parsing and extracting data from various document types for knowledge management systems.

🎭 Best For

🏷️ Topics & Ecosystem

deep-learning document-conversion document-parsing docx epub llm markdown ocr pdf pdf-converter pdf-parser pdf-to-markdown pptx python table-extraction

📊 Activity

Latest commit: 2026-11-18. Activity data is based on daily RepoPi snapshots of the GitHub repository.