Get Your Documents Ready for Gen AI

Docling converts messy documents into structured data and simplifies downstream document and AI processing by detecting tables, formulas, reading order, OCR, and much more.

pip install docling
Duck with documents

Why Choose Docling?

📑Advanced PDF Understanding

Detect page layout, reading order, table structure, code, formulas, and image classification with state-of-the-art models.

Learn more →

🗂️Multiple Formats

Parse PDF, DOCX, PPTX, XLSX, HTML, Markdown, images, audio files, and more into a unified format.

Learn more →

🤖AI Integrations

Plug-and-play integrations with LangChain, LlamaIndex, Crew AI, Haystack, and other popular frameworks.

Learn more →

🔒Local Execution

Run everything locally for sensitive data and air-gapped environments. Your data stays private.

Learn more →

🔍Extensive OCR

Advanced OCR support for scanned PDFs and images with high accuracy text extraction.

Learn more →

💻Simple CLI

Convert documents directly from your terminal with a simple, powerful command-line interface.

Learn more →

Quick Start

Python
from docling.document_converter import DocumentConverter

source = "https://arxiv.org/pdf/2408.09869"
converter = DocumentConverter()
doc = converter.convert(source).document
print(doc.export_to_markdown())

Supported Document Formats

PDF
Word (DOCX)
PowerPoint (PPTX)
Excel (XLSX)
Markdown
HTML
AsciiDoc
CSV
WebVTT
MP3
WAV
PNG
JPEG
TIFF
BMP
WEBP

And many more! See full list

Export Formats

Export your parsed documents to formats that simplify processing and ingestion into AI, RAG, and agentic systems.

Text

Plain text extraction

Markdown

Structured markdown with tables and formatting

HTML

Rich HTML output with styling

JSON

Lossless JSON representation

DocTags

Structured document tags format

Use Cases

Docling powers document processing across various industries and applications

📚

Research & Academia

Process research papers, extract citations, analyze academic documents, and convert PDFs to structured formats for literature reviews and knowledge extraction.

💼

Legal Document Processing

Extract key information from legal documents, contracts, and case files. Maintain document structure while enabling search and analysis.

📊

Business Intelligence

Convert business reports, financial statements, and presentations into structured data for analysis, reporting, and decision-making.

🏥

Healthcare & Medical Records

Process medical documents, patient records, and research papers while maintaining privacy with local processing capabilities.

📝

Content Management

Convert documents to Markdown or HTML for content management systems, documentation sites, and knowledge bases.

🤖

AI & Machine Learning

Prepare documents for RAG systems, fine-tuning datasets, and AI model training. Extract structured data for machine learning pipelines.

Powered by Advanced Technology

Heron Layout Model

State-of-the-art layout detection model for fast and accurate PDF parsing. Understands complex page structures, multi-column layouts, and reading order.

Visual Language Models

Support for VLMs like GraniteDocling with MLX acceleration on Apple Silicon. Enhanced understanding of document content through vision-language integration.

Advanced OCR

High-accuracy OCR for scanned documents and images. Supports multiple languages and character sets with intelligent text extraction.

Table Detection

Intelligent table extraction preserving cell relationships, headers, and data structure. Export tables in multiple formats for downstream processing.

Performance & Scalability

Fast Processing

Optimized pipelines for quick document conversion. Process documents in seconds, not minutes.

Batch Processing

Handle multiple documents efficiently with parallel processing support. Scale from single documents to large document collections.

Memory Efficient

Stream large documents without loading everything into memory. Process documents of any size efficiently.

Cross-Platform

Works seamlessly on macOS, Linux, and Windows. Supports both x86_64 and arm64 architectures including Apple Silicon.

Join the Community

Docling is part of the LF AI & Data Foundation and backed by IBM Research. Join thousands of developers building the future of document processing.

Open Source

MIT licensed and open source. Contribute, fork, and use freely in your projects.

Active Development

Regular updates with new features, improvements, and bug fixes. Active community support and contributions.

Enterprise Ready

Used by enterprises worldwide. Production-ready with comprehensive documentation and support.

Ready to Get Started?

Join thousands of developers using Docling to process documents for AI applications.