Geospatial platform as a service

Maps, imagery and analysis for every city, on one API

Atlas gives you the basemap, places, routing, satellite imagery and an Earth Engine-compatible analysis runtime in a single subscription. Build a city app, model urban heat, or feed a digital twin, without stitching five vendors together.
Vector, raster, terrain and imagery tiles
Runs Earth Engine scripts unmodified
3D Tiles, OpenDRIVE and OpenSCENARIO export
One platform

From the street to the satellite

Most teams buy a map vendor, a geocoder, a routing engine and a remote-sensing platform, then glue them together. Atlas is all four, sharing one data model.
01

Map it

Serve your own basemap from Overture data: vector tiles for MapLibre, rendered raster and terrain tiles, places and addresses.
02

Move through it

Drive, walk, bike and transit routing, isochrones and travel-time comparisons on a road graph Atlas builds and owns.
03

Measure it

Overlay Landsat, Sentinel-2, MODIS and your own rasters, and run Earth Engine-style analysis over just the pixels you ask for.
Capabilities

One API for apps, analysis and simulation

The pieces a geospatial product needs, sharing one data model and one token.

Maps and tiles

Vector tiles (MVT) of roads, buildings, land use, water and places, plus rendered raster and shaded-terrain tiles, in standard {z}/{x}/{y} paths.
Learn more

Places and geocoding

Full-text place search, forward and reverse geocoding, nearest-by-category and walkability facets over Overture places and addresses.
Learn more

Routing and transit

Turn-by-turn directions, isochrones, one-call travel-time comparisons and GTFS trip planning, on Atlas’s own routing graph.
Learn more

Satellite imagery

Ingested remote-sensing datasets served as tiles straight from cloud-optimized GeoTIFFs, with colormaps, rescale ranges and band algorithms per request.
Learn more

Earth Engine-compatible analysis

Paste a Code Editor script and get the same maps, printed numbers and charts, computed by Atlas over public archives and your own data.
Learn more

Shareable reports

Every script run saves a report: the script, its output, each chart’s data and every map layer, re-rendered on demand and exportable to kepler.gl.
Learn more

Digital-twin export

Export a region as OGC 3D Tiles, ASAM OpenDRIVE and OpenSCENARIO, ready for simulators such as CARLA.
Learn more

Boundaries and conditions

Administrative divisions, elevation, live weather observations and road hazards for any covered point.
Learn more

Custom route ranking

Rank routes by your own metrics, not just travel time: send a grid of anything you can map (heat, noise, air quality) with a weight, and Atlas finds the best route under it.
Learn more
Tour the platform
Earth Engine compatibility

Bring your Earth Engine scripts. Keep the code.

Atlas runs Google Earth Engine Code Editor scripts as they are, with the ee, Map, ui, print and Export globals, and gives you the same maps, numbers and charts, computed on Atlas.
// Unchanged Earth Engine Code Editor script var roi = ee.Geometry.Rectangle([126.76, 37.42, 127.18, 37.70]); // Seoul var lst = ee.ImageCollection('MODIS/061/MOD11A2') .filterDate('2025-07-01', '2025-08-31') .select('LST_Day_1km') .mean() .multiply(0.02).subtract(273.15); print('Mean summer LST (°C)', lst.reduceRegion(ee.Reducer.mean(), roi, 1000)); Map.addLayer(lst, {min: 25, max: 45, palette: ['blue', 'yellow', 'red']}, 'LST');
The script is the same one you’d paste into the Code Editor.
Compare

The work of five platforms, in one

Atlas overlaps with each of these, and replaces the glue between them.
Atlas vs Google Earth Engine
Runs the same Code Editor scripts on Atlas’s own compute, scoped to the pixels each request needs, and pairs the results with a basemap, routing and places. Atlas doesn’t replace planet-wide batch processing.
Atlas vs Mapbox
Covers the same map, search and routing jobs from an open data base (Overture), and adds satellite imagery, analysis and simulation export on the same API.
Atlas vs CARTO
Hosts the data and the compute itself, so there is no warehouse to set up, and covers raster imagery and routing alongside vector data.
Atlas vs Microsoft Planetary Computer
Reads those same public archives over STAC and adds the analysis runtime, tiles and app-ready APIs on top, so you don’t assemble the compute yourself.
Atlas vs Planet Labs
Doesn’t operate satellites. Atlas analyzes public archives and any imagery you ingest as cloud-optimized GeoTIFFs, then puts it in context with streets, buildings and routing.
Full comparison

Put a whole city behind one API

Make your first request in a few minutes: search a place, route to it and pull the imagery around it.