docs/usage/api_server/index.md
!!! info "Synced from docling-serve v1.21.0" This page summarizes the docling-serve documentation at v1.21.0. For the exhaustive reference, follow the links to the source repository.
Run Docling as an HTTP service with docling-serve — a FastAPI server that exposes Docling's document conversion over a REST API.
| You want to… | Use |
|---|---|
| Call Docling over HTTP from any language (including Python), or share one conversion service | the API server — self-host it, or use a managed service |
| Run Docling directly (in-process) in a Python application | the Python library |
| Run large-scale or distributed batch conversions | Jobkit |
| Expose Docling as tools to an AI agent | the MCP server |
!!! note "Using the API server from an MCP agent"
The MCP server can convert through this REST API instead of running Docling locally — set DOCLING_CONVERSION_MODE=remote and point it at your service URL. See MCP server.
To use the API server you need a running endpoint — run your own or use a managed service — plus its service URL and, if the server requires one (DOCLING_SERVE_API_KEY), an API key. A local server defaults to http://localhost:5001 (interactive API docs at /docs):
curl -X POST "http://localhost:5001/v1/convert/source/async" \
-H "Content-Type: application/json" \
-d '{"http_sources": [{"url": "https://arxiv.org/pdf/2501.17887"}]}'
See Deployment to run and configure it, and REST API for the full API.
A request hits the docling-serve API, which runs the conversion through Docling and returns the result (synchronously, or as an async task you poll). Background jobs run on a pluggable compute engine — in-process by default, or a Redis-backed queue for scaling. For the full API see REST API.