Is an API key required for conversion?

While the core tool can run locally, using the LLM enhancement features requires an API key for Claude, OpenAI, or Gemini, or a local Ollama setup.

What document formats can Marker convert?

Marker supports a wide range of formats including PDF, EPUB, PPTX, DOCX, XLSX, HTML, and common image files.

Does it handle mathematical formulas and tables?

Yes, Marker uses LLM enhancement to accurately convert complex tables into Markdown/JSON and mathematical equations into LaTeX format.

Can I use this for RAG pipelines?

Yes, Marker includes a specific 'chunks' output format that generates flattened JSON optimized for ingestion into vector databases.

Can I extract images from documents?

Yes, image extraction is enabled by default. You can also choose to disable it if you only need the plain text content.

Marker Document Converter

Name: Marker Document Converter
Author: xdanger

byxdanger

0•

コンテンツ管理

Converts PDFs and complex documents into clean, structured Markdown or JSON using AI-powered extraction for tables and math.

Marker is a high-performance document conversion skill that transforms PDFs, EPUBs, and Office files into clean Markdown, HTML, or JSON. By leveraging multimodal LLMs like Claude Haiku and Sonnet, it excels at accurately extracting complex tables, mathematical equations in LaTeX, and multi-column layouts that traditional OCR tools often struggle with. This skill is particularly valuable for developers building RAG (Retrieval-Augmented Generation) pipelines, as it provides optimized data chunks ready for vector database ingestion and handles everything from scanned documents to handwriting.

主な機能

01AI-enhanced extraction for tables, forms, and LaTeX mathematical equations

02Support for PDF, EPUB, PPTX, DOCX, XLSX, HTML, and image formats

03Multiple output modes including Markdown, HTML, JSON, and RAG-optimized chunks

04Integration with Claude, GPT-4o, Google Gemini, and local Ollama models

050 GitHub stars

06Configurable OCR processing for scanned documents and handwriting

ユースケース

01Converting complex research papers and technical manuals into clean Markdown for documentation

02Preparing large document sets for RAG pipelines and vector database ingestion

03Extracting structured data from financial reports or tables into JSON format

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add xdanger/claude-plugins marker

For use in Claude.ai and ChatGPT

Download Skill