skills.rest
staruhub
staruhubCOMMUNITY
Explore all skills by this author →
GitHub

mineru-pdf-parser

Convert complex PDF documents into Markdown and JSON using MinerU.

706|130|Updated Oct 31, 2025
Search Tags:#document processing#OCR#table extraction#Markdown conversion#MinerU#PDF parsing#rAG
One-click install
npx skills add https://github.com/staruhub/ClaudeSkills --skill mineru-pdf-parser
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: mineru-pdf-parser
Source: https://github.com/staruhub/ClaudeSkills/tree/main/skills/Geek-skills-mineru-pdf-parser
Command: npx skills add https://github.com/staruhub/ClaudeSkills --skill mineru-pdf-parser

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires mineru, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Complex PDFs with formulas, tables, and multi-column layouts are hard to feed into LLMs or RAG pipelines. This Skill converts them into clean Markdown and structured JSON so the content becomes machine-readable without manual copy-paste or reformatting.

Core Features & Use Cases

  • PDF to Markdown/JSON Conversion: Parse PDFs with three backends (pipeline, vlm, hybrid) and output Markdown, structured content-list JSON, and extracted images.
  • Batch Processing: Parse entire directories of PDFs concurrently with per-file success/failure reporting via the included script.
  • RAG Data Preparation: Extract sections, tables, and formulas as structured chunks ready for embedding and vector database ingestion.
  • Use Case: You have a folder of academic papers with dense LaTeX formulas. Run the parser in vlm mode to get Markdown with intact formulas and tables, then load the section chunks into your vector database.

Quick Start

Use the mineru-pdf-parser skill to convert paper.pdf into Markdown and JSON in the output folder using the vlm backend.

Frequently Asked Questions about mineru-pdf-parser

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I convert a PDF to Markdown with MinerU?▼

Run mineru -p input.pdf -o output_dir after installing with uv pip install -U "mineru[all]" and downloading models via mineru-models-download. The output directory contains a .md file, a content_list JSON, and extracted images.

Which MinerU backend should I use for PDF parsing?▼

Use pipeline for simple text PDFs since it is fast and lightweight, vlm for academic papers with dense formulas and tables since it has the highest accuracy, and hybrid when you are unsure since it balances speed and precision.

Can MinerU extract tables and formulas from PDFs?▼

Yes, MinerU extracts tables as Markdown or HTML and formulas as LaTeX, along with images saved to an images folder. The vlm backend gives the best results for formula- and table-heavy documents.

Why is MinerU vlm mode extremely slow on my machine?▼

VLM mode is very slow on CPU and can appear frozen. Without a GPU, switch to pipeline or hybrid backend; with a GPU, set device-mode to cuda in ~/.mineru.json to enable acceleration.

What are the limitations of MinerU for scanned PDFs?▼

Low-resolution scanned PDFs produce poor parsing quality since results depend on scan clarity. Specify the OCR language with --lang, retry with a different backend, and report honestly if quality remains unacceptable.

How do I prepare PDF content for a RAG pipeline?▼

Parse the PDF with MinerU, call result.get_sections() to obtain titled content chunks with page metadata, then embed each chunk and store it in your vector database. The content_list JSON also works for structured chunking.

Related Skills

Explore More Data & Analytics Skills→
syzl8512syzl8512

12306-invoice-downloader

Download and rename 12306 train ticket invoice attachments from 163 email.

Community
Intermediate
jackjin1997jackjin1997
11

LangChain RAG Pipeline

Build end-to-end RAG pipelines with document loading, chunking, and vector storage.

Community
Intermediate
neverpreparedneverprepared

pdf-harvester

Extract text, tables, and metadata from PDFs for RAG workflows.

Official
Advanced
decocmsdecocms
396

pdf

Extract embedded text from PDF files page by page.

Official
Basic