Docling

What is Docling?

Docling is an open-source library that turns messy documents into structured data that AI applications can use. It reads a long list of formats: Apple Pages, Numbers and Keynote files, EPUB, HTML, XLSX, PPTX, DOCX and PDF, plus images, LaTeX, email files, and audio or video through speech recognition. For PDFs it works out page layout, reading order, table structure, code, formulas and image classes, and it applies OCR to scans. Everything lands in one unified document format that exports to Markdown, HTML, WebVTT or lossless JSON. You can run it from a Python script, a command line, an MCP server or the docling-serve API server, and it can execute locally for sensitive or air-gapped data. Integrations cover LangChain, LlamaIndex, Crew AI and Haystack. Docling is free under the MIT license, started at IBM Research Zurich and is hosted by the LF AI and Data Foundation.

Use cases

• Convert PDFs to Markdown
• Prepare documents for RAG pipelines
• Extract tables from scanned files
• Parse Word and PowerPoint files
• Transcribe audio with ASR
• Export documents as lossless JSON
• Feed documents to AI agents over MCP

Who it's for

• Python developers
• Data engineers
• AI engineers
• Researchers

Why use Docling

• It is free and MIT licensed, so there is no per-page fee or vendor quota.
• Documents can be processed locally, which suits sensitive data and air-gapped setups.
• One document model feeds Markdown, HTML and JSON outputs from many source formats.
• Ready-made integrations save glue code in LangChain, LlamaIndex, Crew AI and Haystack projects.
• The CLI, Python API, MCP server and API server fit scripts, services and agent tooling alike.

Docling features

• Parses PDF, DOCX, PPTX, XLSX, HTML, EPUB and image files
• PDF understanding of layout, reading order, tables, code and formulas
• OCR for scanned PDFs and images
• Unified DoclingDocument format
• Export to Markdown, HTML, WebVTT, DocTags and JSON
• GraniteDocling and other visual language models
• Audio transcription with speech recognition models
• MCP server and docling-serve API server
• Local execution for air-gapped environments
• Command-line interface

Docling integrations

• LangChain
• LlamaIndex
• Crew AI
• Haystack
• Hugging Face
• Apify

How to get started with Docling

1. Install the package with pip install docling on macOS, Linux or Windows.
2. Run the docling command on a local file or URL, or call DocumentConverter from Python.
3. Docling parses the layout, tables and text, applying OCR or a vision model where needed.
4. Export the result as Markdown or JSON and pass it to your retrieval or agent pipeline.

Docling limitations

• Python 3.9 support was dropped in version 2.70.0, so you need Python 3.10 or higher
• Individual models carry their own licenses outside the MIT-licensed codebase
• Metadata extraction and chemistry understanding are listed as coming soon
• You run it on your own hardware

Docling pricing

Docling is open source, so you can use and self-host it for free.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

Docling alternatives

The main Docling alternatives listed on Debuly are Unstructured, Reducto, Mathpix, Crawl4AI, Expand and Firecrawl. The table compares pricing and platforms; order is A–Z.

1. 

   Crawl4AI

   Open-source web crawler and scraper that turns pages into clean Markdown for LLMs and RAG, with an optional hosted cloud API.

   Developer platformsResearch and searchAutomation and agents
   FreemiumVisit Crawl4AI (official site)
2. 

   Expand

   Web data API that fetches pages, files and videos as agent-ready Markdown, hidden JSON and citable snapshots

   Developer platformsResearch and searchAutomation and agents
   FreemiumVisit Expand (official site)
3. 

   Firecrawl

   Web data API for AI agents: search, scrape, crawl and interact with websites and get clean content back, with a free tier.

   Developer platformsResearch and searchAutomation and agents
   FreemiumVisit Firecrawl (official site)
4. 

   Mathpix

   OCR software and API that turns screenshots, scans, handwriting and PDFs into LaTeX, Markdown, DOCX and other editable formats.

   ProductivityDeveloper platformsWriting
   FreemiumVisit Mathpix (official site)
5. 

   Reducto

   Document processing API and Studio that parses PDFs, scans and spreadsheets into structured data for AI and data teams.

   Developer platformsAutomation and agents
   Free trialVisit Reducto (official site)
6. 

   Unstructured

   Turns PDFs, Office files, emails and images into clean, chunked, embedded data that is ready for LLM and RAG pipelines.

   Developer platformsAutomation and agentsProductivity
   FreemiumVisit Unstructured (official site)

Docling FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from avatars.githubusercontent.com · fetched 24 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.