---
title: "Deploy Docling"
description: "Convert PDF, Office files and scans to Markdown/JSON. API key, web UI"
category: "AI/ML"
url: https://railway.com/deploy/docling-1
---

# Deploy Docling

Convert PDF, Office files and scans to Markdown/JSON. API key, web UI

**[Deploy Docling on Railway](https://railway.com/template/docling-1)**

Machine-readable deploy manifest (JSON, validated by TemplateCI): https://railway.com/deploy/docling-1/manifest.json

- **Creator:** Agaz Self-Host
- **Category:** AI/ML
- **Total deploys:** 2

## Template content

### Docling https://cdn.jsdelivr.net/gh/homarr-labs/dashboard-icons/png/docling.png

- **Source:** https://github.com/aalfath/docling-railway-template
- **Health check:** /ready
- **Public domain:** Yes

## Documentation

# Deploy and Host Docling on Railway

Docling is an open-source document conversion toolkit from the Docling project (LF AI &amp; Data). It reads PDF, DOCX, PPTX, XLSX, HTML, images and scanned pages, works out layout, reading order and tables, runs OCR where needed, and exports clean Markdown, JSON, HTML or text. Docling Serve wraps it in an HTTP API and web UI.

## About Hosting Docling

This template runs Docling Serve v1.38.2, the CPU build of the official image, as one public service with a Railway domain. The layout, table, picture-classifier and OCR models ship inside the image, so the service needs no downloads at boot and reports ready about 15 seconds after it starts. Every API call needs the `X-Api-Key` header. The key is generated at deploy time, and the service refuses to start without one. Docling Serve is stateless, so there's no database or volume. Plan for 2 to 4 GB of RAM, more for large documents. The first build pulls a 4.4 GB image and takes a few minutes.

## Common Use Cases

* Turn PDFs, Office files and scans into Markdown or JSON for RAG pipelines and vector databases
* Give LLM agents and automation tools (n8n, LangChain, LlamaIndex) a private document parsing API
* Extract tables and structured text from reports, invoices and papers without sending files to a third-party service

## Dependencies for Docling Hosting

* Docling Serve container image (CPU build), built from a public GitHub repo
* About 2 to 4 GB of RAM

### Deployment Dependencies

* Docling Serve: https://github.com/docling-project/docling-serve
* Release v1.38.2: https://github.com/docling-project/docling-serve/releases/tag/v1.38.2
* Configuration reference: https://github.com/docling-project/docling-serve/blob/main/docs/configuration.md
* Template repo: https://github.com/aalfath/docling-railway-template

### Implementation Details

Open `https://YOUR-DOMAIN/` for the web UI and paste the API key (Variables tab, `DOCLING_SERVE_API_KEY`) into its server panel. API docs are at `/docs`.

```sh
curl -X POST https://YOUR-DOMAIN/v1/convert/file \
  -H "X-Api-Key: YOUR-API-KEY" \
  -F "files=@document.pdf" -F "to_formats=md"
```

Other services in the project can call `http://${{Docling.RAILWAY_PRIVATE_DOMAIN}}:5001` over the private network. Any Docling Serve setting can be added as a variable, for example `DOCLING_SERVE_MAX_NUM_PAGES` or `UVICORN_WORKERS`. Updates arrive as commits to the template repo, and Railway shows an "update available" notice when one lands.

## Why Deploy Docling on Railway?


Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying Docling on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.



## Similar templates

- [Chat Chat](https://railway.com/deploy/-WWW5r) — Chat Chat, your own unified chat and search to AI platform.
- [Weaviate](https://railway.com/deploy/weaviate-1) — Vector database with hybrid search and an API key-protected API
- [Möbius Personal AI OS](https://railway.com/deploy/mobius-v01-personal-ai-os) — Self-hosted AI agent that builds and runs apps for you, with memory.

Open this page in a browser: https://railway.com/deploy/docling-1
