molandev/markitdown-converter-api repository overview

⁠Document Converter Service - Overview

⁠📦 Source Code

• GitHub: https://github.com/molandev/markdown-converter⁠
• Gitee: https://gitee.com/molandev/markdown-converter⁠

⁠⚠️ Important Notice

⁠🚀 What Is It?

A lightning-fast, cloud-native document conversion service powered by GraalVM Native Image and Python's MarkItDown. Built for high-performance scenarios with minimal resource consumption, it transforms Office documents into clean Markdown format effortlessly.

⁠✨ Core Features

⁠🔥 Blazing Fast Performance

• GraalVM Native Image: Sub-second startup, 50% less memory footprint
• MarkItDown Powered: Intelligent extraction from Office documents
• Stream Processing: Handles large files (100MB+) without OOM errors

⁠🛡️ Production-Ready Reliability

• Smart Concurrency Control: Semaphore-based rate limiting prevents system overload
• Auto-Cleanup Mechanism: Scheduled task automatically purges temporary files
• Graceful Degradation: Queuing mechanism with configurable timeouts

⁠🎯 Developer-Friendly API

• Dual Input Modes: Upload files directly or provide URLs
• RESTful Design: Clean endpoints with proper HTTP status codes
• Chinese Filename Support: URL-encoded headers for international compatibility

⁠📦 Supported Formats

⁠🎮 Quick Start

⁠Pull from Docker Hub

docker pull molandev/markitdown-converter-api:latest

⁠Run in Seconds

docker run -d \
  --name doc-converter \
  -p 10996:10996 \
  -v /data/temp:/app/temp \
  -e CONVERTER_MAX_CONCURRENT=10 \
  -e CONVERTER_RETENTION_MINUTES=10 \
  molandev/markitdown-converter-api:latest

⁠📚 API Documentation

Access the complete API guide at: http://localhost:10996/help

⁠Convert Your First Document

# Upload a Word document
curl -F "[email protected]" \
     http://localhost:10996/convert/markitdown/upload \
     -o report.md

# Or provide a URL
curl -X POST "http://localhost:10996/convert/markitdown/url?fileUrl=https://example.com/doc.pdf" \
     -o output.md

# Keep embedded images
curl -F "[email protected]" \
     -F "keepDataUris=true" \
     http://localhost:10996/convert/markitdown/upload \
     -o output.md

⁠⚙️ Configuration

⁠🔧 Advanced Usage

⁠High-Concurrency Setup

docker run -e CONVERTER_MAX_CONCURRENT=20 \
           -e CONVERTER_ACQUIRE_TIMEOUT=60 \
           molandev/markitdown-converter-api:latest

⁠Minimal Resource Environment

docker run -e CONVERTER_MAX_CONCURRENT=3 \
           -e CONVERTER_RETENTION_MINUTES=5 \
           molandev/markitdown-converter-api:latest

⁠💡 Why Choose This?

✅ Zero JVM Warmup - GraalVM Native means instant peak performance
✅ Resource-Efficient - Runs comfortably on 2C4G instances
✅ Production-Ready - Smart concurrency control & auto-cleanup
✅ Cloud-Native - Docker-first design, Kubernetes-ready
✅ MarkItDown Powered - Best-in-class Office document extraction

⁠🎯 Perfect For

• AI Knowledge Base Pipelines: Convert Office docs to LLM-friendly Markdown
• Document Management Systems: Batch conversion workflows
• Content Migration Projects: Legacy format modernization
• SaaS Applications: Multi-tenant document processing

Built with Spring Boot 3.5.6 + GraalVM 25 + Python MarkItDown

One engine, one purpose: Perfect Markdown from Office documents. ⚡