docling

Repository

Extracts text, tables, and layout from documents and scans for export to Markdown, HTML, or JSON.

Overview

Use case: Source-backed content preparation

A document processing toolkit parses PDFs and other formats into a unified representation. It handles page layout, reading order, tables, and scanned text, with exports including Markdown, HTML, and JSON.

Developers can convert local files or URLs using Python or the command line, then connect results to AI frameworks. Local execution also supports sensitive documents and environments without network access, helping prepare varied sources for AI applications.

Source-based overview · Not hands-on tested.

Good for

Extract tables from PDF documents
Read scanned documents with OCR
Export document content as Markdown
Convert documents to structured JSON
Process sensitive documents locally

Price

Not stated

Setup

Technical setup (inferred)
Install the Python package with pip and use Python 3.10 or higher. Conversion is available through Python or the CLI on macOS, Linux, and Windows.

Change history

No substantive changes recorded. A successful source check alone is not a change.

Links work · Not hands-on tested · Recently active: source activity within 14 days · Updated: recorded change this week

Compare all candidates →

Licence, sources & verification

Listing facts

MIT

Listing facts
Source	https://github.com/docling-project/docling
Author credit	docling-project on GitHub
Licence	MIT
Use case / type	Source-backed content preparation · repository

Open action link

Evidence

Source-discovered · Not hands-on tested · Links not checked

Source activity date: 22 Oct 2026
Checked: 22 Oct 2026 · Imported from a source snapshot; not independently rechecked
Source-metadata observation: Not recorded

Source: Not checked; availability unconfirmed

Workflow context

Intended task: Prepare a draft, inspect its claims, and keep the receipts.
Setup: Technical setup (inferred)
Install the Python package with pip and use Python 3.10 or higher. Conversion is available through Python or the CLI on macOS, Linux, and Windows.
Prerequisites: Review repository setup instructions, runtime, model access and API credentials; exact requirements are not established here.
Limitations: Source checking does not prove functionality. No hands-on test was performed. Capabilities and availability are unconfirmed.
Price: Not stated

Attribution & reuse

Source: docling-project on GitHub.

Repository licence: MIT (check the repository for scope).

Changes made: summary paraphrased from the repository description with AI assistance, plus task tags; no README text, code or media reproduced.

Recorded modifications: summary written by TaskToTool in its own words from the repository's GitHub description; reviewed against the live source on 2026-10-22.

Editorial provenance

Summary written by TaskToTool from public source information; not hands-on tested.

Gets documents into shape for use with generative AI.

Not hands-on tested.

AI-assisted drafting; reviewed by TaskToTool

Reviewed by TaskToTool Content Editor · 22 Oct 2026

How we write and verify listings: AI assistance, evidence labels and dates