# Docling review

> Docling is a Python SDK and CLI that parses PDF, DOCX, HTML and other formats into a unified document representation for gen AI.

Devtoolwatch Board, the week of Oct 25–31, 2026. Canonical page: https://index.devtoolwatch.com/tools/docling/

## TL;DR

Docling is a Python library and CLI that converts many document, audio and video formats into a unified DoclingDocument format and exports Markdown, HTML or lossless JSON. Docling is for developers preparing documents for RAG and agent pipelines. It adds advanced PDF understanding, OCR, VLM support and local execution for sensitive data.

## Docling at a glance

- **Category:** [RAG, data and memory](https://index.devtoolwatch.com/categories/rag-data-and-memory/)
- **Maintained by:** Docling Project (organization, [@docling-project](https://github.com/docling-project) on GitHub)
- **License:** MIT
- **Language:** Python, Shell, Dockerfile
- **Repository:** [github.com/docling-project/docling](https://github.com/docling-project/docling)
- **Website:** [docling-project.github.io/docling](https://docling-project.github.io/docling)
- **PyPI package:** [docling](https://pypi.org/project/docling/) 2.141.0
- **Latest release:** [v2.141.0](https://github.com/docling-project/docling/releases/tag/v2.141.0) on GitHub, Oct 20, 2026
- **First published:** Jul 9, 2024

## What Docling does

- Docling parses PDF, DOCX, PPTX, XLSX, HTML, EPUB, Apple iWork files, email, images, LaTeX, audio and video files.
- Docling analyzes PDF page layout, reading order, table structure, code, formulas and image classification.
- Docling exports to Markdown, HTML, WebVTT, DocLang, DocTags and lossless JSON.
- Docling provides OCR for scanned PDFs and images, and speech recognition for audio files.
- Docling supports Visual Language Models such as GraniteDocling, and v2.141.0 added nvidia/Nemotron-3.5-Parse.
- Docling converts bar charts, pie charts and line plots into tables or code with descriptions.
- Docling integrates with LangChain, LlamaIndex, Crew AI and Haystack, and offers an MCP server for agents.

### How it works

Docling installs as a Python package. You use it through the DocumentConverter class in Python, which the project recommends, or through the docling CLI. Docling can also run as a service through the docling-serve API server, or connect to agents through its MCP server. Docling can run locally, including in air-gapped environments, and plugs into RAG frameworks and vector stores.

## How to install Docling

**With pip**

```
pip install docling
```

### Requirements

- Python 3.10 or higher, below 4.0. Python 3.9 support was dropped in version 2.70.0.
- Runs on macOS, Linux and Windows, on x86_64 and arm64.

### First steps

1. Install Docling with pip on Python 3.10 or higher.
2. Convert a document from the CLI with `docling https://arxiv.org/pdf/2206.01062`. This writes a .md file to the current directory.
3. In Python, create a DocumentConverter, call convert() on a local path or URL, then call export_to_markdown().
4. To use a VLM, run `docling --pipeline vlm --vlm-model granite_docling https://arxiv.org/pdf/2206.01062`.

## Who Docling is for

- Developers building RAG pipelines that need structured text from PDFs and office documents.
- Teams processing sensitive documents that must stay local or air-gapped.
- Agent builders who want document parsing through an MCP server or REST API.
- Users working with domain formats such as USPTO patents, JATS articles or XBRL financial reports.

### Limitations

- Docling requires Python 3.10 or newer, so older Python environments cannot use current versions.
- Docling's codebase is MIT licensed, but individual models carry their own licenses, which you must check separately.
- Docling lists metadata extraction and chemistry understanding as coming soon, so they are not yet available.

## Usage data (the week of Oct 25–31, 2026)

- Board rank: #125 overall, #12 in RAG, data and memory, momentum score 57
- Downloads that week: 904.2k (+8% vs 4-week average)
- New GitHub stars per week (last 12 weeks): Aug 9: 398; Aug 16: 651; Aug 23: 345; Aug 30: 340; Sep 6: 290; Sep 13: 929; Sep 20: 798; Sep 27: 373; Oct 4: 344; Oct 11: 415; Oct 18: 531; Oct 25: 425

## Docling alternatives

| Rank | Tool | Stars | Downloads/wk |
|---|---|---|---|
| #1 | [XERJ](https://index.devtoolwatch.com/tools/xerj-org-xerj/) | 17.9k | – |
| #2 | [Memvid](https://index.devtoolwatch.com/tools/memvid-memvid/) | 17.3k | – |
| #3 | [LEANN](https://index.devtoolwatch.com/tools/startrail-org-leann/) | 13.7k | – |
| #4 | [Graphify](https://index.devtoolwatch.com/tools/graphify-labs-graphify/) | 140k | 480.7k |
| #5 | [MarkItDown](https://index.devtoolwatch.com/tools/markitdown/) | 193.2k | 4.7M |
| #6 | [Firecrawl](https://index.devtoolwatch.com/tools/firecrawl/) | 197.4k | 2.9M |

## Docling FAQ

### Is Docling free and open source?

Yes. Docling's source code is on GitHub (docling-project/docling) under the MIT license, so you can use it for free.

### How do I install Docling?

Run `pip install docling`. The docling package needs Python 3.10 or newer, below 4.0.

### Is Docling still maintained?

Yes. The docling-project/docling repository was last updated on Oct 21, 2026, and the latest release, v2.141.0 on GitHub, came out on Oct 20, 2026.

### How popular is Docling?

Docling has 69.9k GitHub stars as of Nov 1, 2026. It gained 425 new stars in the week of Oct 25–31, 2026, against a 12-week average of 487 a week. Its main package had 904.2k downloads that week, +8% against its own 4-week average. That made it #12 among RAG, data and memory tools on the Devtoolwatch Board.

### Docling vs MarkItDown: how do they compare?

Docling and MarkItDown are both tracked by Devtoolwatch as RAG, data and memory tools. Docling has 69.9k GitHub stars and MarkItDown has 193.2k. In the week of Oct 25–31, 2026, Docling's main package had 904.2k downloads and MarkItDown's had 4.7M. MarkItDown is a lightweight Python utility from Microsoft that converts PDFs, Office documents, images, audio and other files to Markdown for LLMs. [MarkItDown review](https://index.devtoolwatch.com/tools/markitdown/)

### Docling vs MinerU: how do they compare?

Docling and MinerU are both tracked by Devtoolwatch as RAG, data and memory tools. Docling has 69.9k GitHub stars and MinerU has 82.9k. In the week of Oct 25–31, 2026, Docling's main package had 904.2k downloads and MinerU's had 30.8k. MinerU is a Python document parser that converts PDFs, images, and Office files into Markdown or JSON for LLM and agent workflows. [MinerU review](https://index.devtoolwatch.com/tools/mineru/)

### What are the best alternatives to Docling?

Other RAG, data and memory tools on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: XERJ (#1), Memvid (#2), LEANN (#3), Graphify (#4), MarkItDown (#5). The alternatives table above has their numbers.

### How do I use Docling in Python?

Docling's Python usage centers on DocumentConverter. Import it from docling.document_converter, create a converter, call convert() with a local path or URL, then call result.document.export_to_markdown(). The README recommends this approach.

### How do I convert a PDF to Markdown with Docling?

Docling's CLI converts a PDF to Markdown with one command, `docling https://arxiv.org/pdf/2206.01062`. It writes a .md file with the structured content to the current directory.

### Can Docling run locally or offline?

Yes. Docling supports local execution for sensitive data and air-gapped environments.

### Is Docling an OCR tool?

Docling includes OCR but does more than OCR. Docling offers OCR support for scanned PDFs and images, alongside layout analysis, table structure recognition, VLM pipelines and speech recognition for audio.

### How do I use Docling for RAG?

Docling has integrations with LangChain, LlamaIndex, Crew AI and Haystack. The docs include chunking guides and RAG examples with Milvus, Weaviate, Qdrant, OpenSearch and others.

### How do I run Docling as a service?

Docling runs as a service through docling-serve, its API server. The docs cover deployment and a REST API. Docling also offers an MCP server for connecting agents.

## Sources

1. [docling-project/docling on GitHub](https://github.com/docling-project/docling)
2. [Docling README](https://github.com/docling-project/docling/blob/main/README.md)
3. [docling on PyPI](https://pypi.org/project/docling/)
4. [Docling website](https://docling-project.github.io/docling)
5. [Release v2.141.0 (Oct 20, 2026)](https://github.com/docling-project/docling/releases/tag/v2.141.0)
6. [Release v2.140.0 (Oct 16, 2026)](https://github.com/docling-project/docling/releases/tag/v2.140.0)
7. [Release v2.139.0 (Oct 14, 2026)](https://github.com/docling-project/docling/releases/tag/v2.139.0)
8. [Devtoolwatch Board, the week of Oct 25–31, 2026](https://index.devtoolwatch.com/)