Five alternatives to the Google Maps API

Why people switch

Three reasons come up. Per-call billing, which becomes unpredictable as soon as traffic rises. The terms of use, which forbid certain uses: storing results, displaying them outside their map. And the data question: every display sends information to a third party, which is a problem in certain regulated contexts.

If none of the three applies to you, stay where you are. Migrating costs time for a result the user won't see.

The five

MapLibre GL: the display engine, open source and licence-free. It doesn't provide the data: it displays it. It's the building block you pair with one of the sources below.

OpenStreetMap: the data, free and editable. Excellent coverage in urban Europe, more variable elsewhere. Routing and address search require an additional service.

Mapbox: the closest to Google in ease of implementation, with real control over rendering. Still a commercial service billed by usage.

HERE: strong on logistics and heavy goods vehicle routing, with European hosting options.

Swisstopo: official Swiss data, with a precision nobody else matches on Swiss territory: land registry, elevation, historical aerial photographs. For a Swiss project rooted in the terrain, it's often the best source, and it's public.

How to choose

If you display points on a map: MapLibre and OpenStreetMap are enough, and the bill disappears.

If you calculate routes with real constraints: HERE or Mapbox.

If your project concerns Swiss territory (real estate, energy, public authorities, agriculture): start with Swisstopo before looking elsewhere.

What a migration costs

Replacing a map display is quick. What takes time is everything around it: address search, custom markers, routing, and testing the edge cases.

Allow [order of magnitude to be specified per project] for a simple map with address search. The real cost depends mostly on the number of side features accumulated over time.