Skip to content


Folders and files

Last commit message
Last commit date

Latest commit



36 Commits

Repository files navigation


PyPI Changelog Tests License

Datasette plugin for serving MBTiles map tiles


Install this plugin in the same environment as Datasette.

$ datasette install datasette-tiles


You can try this plugin out at


This plugin scans all database files connected to Datasette to see if any of them are valid MBTiles databases.

It can then serve tiles from those databases at the following URL:


An example map for each database demonstrating the configured minimum and maximum zoom for that database can be found at /-/tiles/db-name - this can also be accessed via the table and database action menus for that database.

Visit /-/tiles for an index page of attached valid databases.

You can install the datasette-basemap plugin to get a basemap default set of tiles, handling zoom levels 0 to 6 using OpenStreetMap.

Tile coordinate systems

There are two tile coordinate systems in common use for online maps. The first is used by OpenStreetMap and Google Maps, the second is from a specification called Tile Map Service, or TMS.

Both systems use three components: z/x/y - where z is the zoom level, x is the column and y is the row.

The difference is in the way the y value is counted. OpenStreetMap has y=0 at the top. TMS has y=0 at the bottom.

An illustrative example: at zoom level 2 the map is divided into 16 total tiles. The OpenStreetMap scheme numbers them like so:

0/0  1/0  2/0  3/0
0/1  1/1  2/1  3/1
0/2  1/2  2/2  3/2
0/3  1/3  2/3  3/3

The TMS scheme looks like this:

0/3  1/3  2/3  3/3
0/2  1/2  2/2  3/2
0/1  1/1  2/1  3/1
0/0  1/0  2/0  3/0

datasette-tiles can serve tiles using either of these standards. For the OpenStreetMap / Google Maps 0-at-the-top system, use the following URL:


For the TMS 0-at-the-bottom system, use this:


Configuring a Leaflet tile layer

The following JavaScript will configure a Leaflet TileLayer for use with this plugin:

var tiles = leaflet.tileLayer("/-/tiles/basemap/{z}/{x}/{y}.png", {
  minZoom: 0,
  maxZoom: 6,
  attribution: "\u00a9 OpenStreetMap contributors"

Tile stacks

datasette-tiles can be configured to serve tiles from multiple attached MBTiles files, searching each database in order for a tile and falling back to the next in line if that tile is not found.

For a demo of this in action, visit and zoom in on Japan. It should start showing Stamen's Toner map of Japan once you get to zoom level 6 and 7.

The /-/tiles-stack/{z}/{x}/{y}.png endpoint provides this feature.

If you start Datasette like this:

datasette world.mbtiles country.mbtiles city1.mbtiles city2.mbtiles

Any requests for a tile from the /-/tiles-stack path will first check the city2 database, than city1, then country, then world.

If you have the datasette-basemap plugin installed it will be given special treatment: the basemap database will always be the last database checked for a tile.

Rather than rely on the order in which databases were attached, you can instead configure an explicit order using the tiles-stack-order plugin setting. Add the following to your metadata.json file:

    "plugins": {
        "datasette-tiles": {
            "tiles-stack-order": ["world", "country"]

You can then run Datasette like this:

datasette -m metadata.json country.mbtiles world.mbtiles

This endpoint serves tiles using the OpenStreetMap / Google Maps coordinate system. To load tiles using the TMS coordinate system use this endpoint instead:


Retina tiles

Retina (double resolution) tiles are supported by datasette-tiles if the MBTiles database file contains 512x512 tile images as opposed to the default of 256x256. JavaScript libraries such as Leaflet will serve these tiles with a fixed 256x256 size, which will cause them to be displayed correctly by capable operating systems.


To set up this plugin locally, first checkout the code. Then create a new virtual environment:

cd datasette-tiles
python3 -mvenv venv
source venv/bin/activate

Or if you are using pipenv:

pipenv shell

Now install the dependencies and tests:

pip install -e '.[test]'

To run the tests: