OSRM Pre-Processed Data Files

The Pipeline Is the Painful Part

Running your own OSRM server isn't the hard part — processing the OSM data into a usable graph is. osrm-extract, osrm-partition, and osrm-customize are CPU- and memory-heavy, take a long time on anything beyond a small extract, and need to be re-run every time you want fresher OSM data. That's a fragile, repetitive chore sitting in front of infrastructure you already know how to run.

You want to stay in control of your OSRM server. You just don't want to run the pipeline. Here's the obvious solution.

Same OSRM Files, None of the Processing

Simple Routing runs the extract/partition/customize pipeline for you and hands you the finished .osrm* files — you keep full ownership of the server they run on.

How It Works

Each dataset is a single tar file. Stream it straight into tar and start osrm-routed — that's the whole pipeline.

# 1. Download and untar (your link is in the dashboard's OSRM tab)
mkdir osrm-data
curl -sL "$DOWNLOAD_URL" | tar -xf - -C osrm-data --strip-components=1

# 2. Run
osrm-routed --algorithm mld osrm-data/car.osrm

On a flaky connection? Download to a file first with curl -C - -o osrm-data.tar "$DOWNLOAD_URL" so an interrupted download can resume, then untar it.

Prefer Docker? Mount the extracted folder into the osrm-backend image and run the same osrm-routed command. Use the OSRM version shown next to the download in your dashboard.

What You Get

Would rather not run OSRM yourself at all?

If you'd rather skip the server too, not just the pipeline, check out our Hosted OSRM API — same OSRM engine, fully managed.