← all releases
Datalab·Jul 20, 2026·v2.0.0·open source
Marker
Marker is a high-performance document intelligence pipeline designed to convert PDFs, images, and various office documents into clean Markdown, JSON, HTML, or chunks. It leverages advanced deep learning models, including the Surya VLM, to handle complex layouts, tables, equations, and multi-language text extraction with high accuracy.
featured on blogOCRDocument IntelligencePDF to MarkdownMachine Learning
on the blogDatalab Launches Marker 2.0: An Open-Source PDF-to-Markdown Tool Built for RAG Pipeline 5 min
overview
Marker is a high-performance document intelligence pipeline designed to convert PDFs, images, and various office documents into clean Markdown, JSON, HTML, or chunks. It leverages advanced deep learning models, including the Surya VLM, to handle complex layouts, tables, equations, and multi-language text extraction with high accuracy.
key features
- 01Converts PDF, image, PPTX, DOCX, XLSX, HTML, and EPUB files
- 02Formats tables, equations, inline math, and code blocks
- 03Extracts and saves images while removing headers/footers
- 04Supports optional LLM-based accuracy boosting
- 05Works on GPU, CPU, or MPS hardware
- 06Multilingual support for global document processing
related productsbrowse all →