Skip to content
Vecaria
Go Pro

GIS File Formats Explained

GIS files fall into three families: vector (features), raster (images) and tiles (packaged map data). This guide explains the most common vector formats and points to the tools that convert them, all running locally in your browser.

GeoJSON is a text format (JSON) for points, lines, polygons and their properties. It is the lingua franca of web maps — Leaflet, Mapbox, D3 and most APIs speak it natively. The specification assumes coordinates are WGS84 longitude/latitude, so if your data is in a projected CRS you need to reproject first. Files are readable in any text editor, but large datasets get bulky.

The ESRI Shapefile is the oldest still-common interchange format. Despite the name it is not one file: a dataset needs at least .shp, .shx and .dbf, and optional .prj, .cpg and indexes. Field names are capped at 10 characters and dates have no time component. Every GIS desktop reads and writes it, which is why it refuses to die.

KML is Google Earth’s format: features with styles, folders and balloons. Use it when the deliverable is Google Earth or Google Maps. KMZ is the same file zipped with its icons and images. KML coordinates are also WGS84.

An OGC standard that stores vector layers, tiles and attributes in a single SQLite file. It solves the Shapefile multi-file problem and is the modern default in QGIS.

A streaming-friendly single-file vector format with a spatial index. It reads fast even for very large datasets, because readers can skip to the region they need.

GeoParquet puts vector features in a Parquet column store — the right choice when the next step is analytics (DuckDB, Pandas, Polars). PMTiles packages raster or vector map tiles into one file that map libraries can serve from static storage.

For imagery and elevation grids see GeoTIFF Viewer, Raster Reproject and Terrain Analysis.