RustGIS and GeoLibre¶

RustGIS and GeoLibre are sibling opengeos projects with the same maintainer, so people often ask how they differ and which to use. This page answers that: what each is built for, where each is stronger, and how to use them together. For RustGIS beside QGIS, ArcGIS Pro and ArcGIS Online as well, see the full comparison.

The short answer¶

GeoLibre is the broader, web-first GIS. It is built with web technology (Tauri, React, MapLibre GL JS, DuckDB-WASM, deck.gl) and runs in a browser tab with nothing installed, as a desktop app, and as native Android and iOS apps. It reaches further into data (databases, cloud catalogs, dozens of open-data portals), 3D (Cesium, 3D Tiles, I3S, Gaussian splats), AI (SAM segmentation, ONNX object detection), sharing (share links, review comments, real-time collaboration) and the Python and R ecosystems (a Jupyter widget and an R package).

RustGIS is the narrower, native one. It is a single program written entirely in Rust that draws its own interface on the GPU, with no browser engine or JavaScript runtime inside, and the same program compiled to WebAssembly for the browser. Its interface is a ribbon laid out in the order of the work, every button is a command you can run from the command line, Python or an AI agent, its plugins run sandboxed, and it goes further in the classic desktop GIS jobs of data design, reports and print layouts.

They are complementary rather than competing. Both run the same 1,000+ GeoLibre and Whitebox geoprocessing tools, both keep your data on your device, and both read and write the same open formats. Choose GeoLibre for reach, choose RustGIS for a compact, scriptable desktop GIS, or use both.

Where GeoLibre is stronger¶

• More platforms. GeoLibre runs as a web app, an installable PWA, a desktop app on Windows, macOS and Linux, native Android and iOS apps, a Docker image, a Chrome extension and a Jupyter widget. RustGIS runs on the desktop and in the browser; it has no mobile app and no touch-first layout.
• Data reach. GeoLibre opens more formats in the app itself, including File Geodatabase, DWG, MapInfo TAB, NetCDF-4/HDF5, Kerchunk and Icechunk, ICESat-2 and GEDI. It connects to SQL Server and Apache Iceberg as well as PostGIS and DuckDB, and its plugins browse dozens of catalogs and portals, from NASA Earthdata and Planet to Hugging Face and about 150 US local agencies. RustGIS reads a focused set natively and the rest through GDAL's programs when they are installed (see Data formats).
• SQL in the browser. GeoLibre runs DuckDB Spatial, PGlite/PostGIS and Apache Sedona inside the browser. RustGIS's SQL pane uses DuckDB on the desktop; its web app has no SQL engine yet.
• Cartography depth. GeoLibre has rule-based and expression renderers, heat maps, clustering, diagrams, blend modes, layer groups, a data-defined label engine, and style interchange with SLD, QGIS QML and Mapbox GL JSON. RustGIS covers the common renderers and collision-free labels, without that depth.
• 3D. GeoLibre has a Cesium globe with terrain, 3D Tiles, I3S, Gaussian splats, CZML and Google Photorealistic 3D Tiles. RustGIS has a globe, 2.5D extrusion and LiDAR point layers.
• AI and imagery models. Besides an AI assistant, GeoLibre has SAM segmentation, in-browser ONNX/YOLO object detection and an object-based image analysis workbench. RustGIS has pixel classification (k-means, minimum distance and a trainable neural network).
• Sharing and collaboration. GeoLibre shares projects through links with roles, expiry and passwords, has review comments anchored to the map, story maps, dashboards, and an experimental real-time collaboration mode. RustGIS shares by writing files: web map pages, packages and static exports.
• Python, R and notebooks. The geolibre Python package is a two-way Jupyter widget with a leafmap-style API, the geolibre R package works in RStudio, Quarto and Shiny, and the notebook panel runs JupyterLite or JupyterLab. RustGIS has a Python package that runs commands headlessly and built-in notebooks of command cells, but no Python console in the app and no R package.
• Opening other projects. GeoLibre imports QGIS (.qgs/.qgz) and ArcGIS Pro (.aprx/.mapx) projects. RustGIS opens neither.
• Editing. GeoLibre edits vertices and writes edits back to their sources (GeoPackage, PostGIS, ArcGIS feature services). RustGIS cannot yet edit the vertices of existing features, and edits stay in the project until exported.
• Languages. GeoLibre's interface comes in 21 languages, including right-to-left Arabic, Hebrew and Persian. RustGIS has 16, none right-to-left.

Where RustGIS is stronger¶

• A native program, not a web view. RustGIS is memory-safe Rust compiled to one native program (18–30 MB installers) with no browser engine, JavaScript or Python inside. Large GeoPackage, GeoParquet, FlatGeobuf and Shapefile files are streamed a view at a time rather than loaded into a browser tab's memory, and nothing depends on a web view's limits.
• Everything is a command. Every ribbon button, shortcut and pane action is a command with an id and JSON parameters, and the same ids run from rustgis-cli, the rustgis Python package, notebooks, models, the Script pane, the MCP server and the AI Assistant. A batch of maps can be rendered in CI with one command line. GeoLibre is driven through its interface, its Python and R packages and its MCP server; it has no general command-line tool.
• Sandboxed plugins. RustGIS plugins are WebAssembly modules that can do only what their declared permissions allow, and the same file runs on the desktop, in rustgis-cli and in the browser. GeoLibre's TypeScript plugins can do much more inside the app (panels, map tools, custom layers), but its plugin API states that they "run as trusted code in the app window" and that its policy "is not a security boundary".
• Data design. RustGIS defines domains, subtypes, attribute rules, contingent values, relationship classes, archiving, topology rules and traverse/COGO in the project itself, the geodatabase design work of ArcGIS Pro. GeoLibre has attribute forms with constraints and honours the domains and subtypes of the ArcGIS feature services it edits, but doesn't define its own.
• Reports and layouts. RustGIS builds grouped reports with statistics, charts and subreports, and layouts with map frames, grids, extent indicators and map series. GeoLibre's print layout and atlas cover the common cases, and it has no report builder.
• A ribbon interface. Tabs follow the order of the work (Home, Data, Style, Edit, Analyze, View, Publish), with Command Search (Ctrl+K) to find any command or tool by name. People coming from ArcGIS Pro often find it familiar. GeoLibre uses menus and panels, with a command palette and beginner, intermediate and advanced interface levels.
• Self-contained. Everything RustGIS does with local data runs inside the program, with no network, server or helper process, and offline areas sync edited features back. GeoLibre works offline too, but some features need a remote host (GDAL export, ONNX detection and story map HTML export among them), and on the desktop its raster tools and some vector tools use an optional Python sidecar.

Side by side¶

Using them together¶

The two read and write the same open formats, so data moves between them without conversion: GeoPackage, GeoParquet, FlatGeobuf, Shapefile, GeoJSON, KML/KMZ, GPX, CSV and cloud-optimised GeoTIFF. Neither opens the other's project file yet; move the data across and restyle it.

They also share their geoprocessing. RustGIS's Analyze › Tools runs the GeoLibre and Whitebox suite (natively through the geolibre program on the desktop, and on WebAssembly in the browser), so a tool learned in one works the same way in the other. See Analysis.

A common split is to explore, share and collaborate in GeoLibre, especially from a browser, a phone or a notebook, and to use RustGIS for repeatable desktop work: a project scripted command by command, maps rendered by rustgis-cli in CI, data with domains and attribute rules, or reports and map series printed to PDF.

Common questions¶

Why are there two? They are built on different foundations for different jobs. GeoLibre uses the web platform to reach as many places and data sources as possible. RustGIS explores what a desktop GIS looks like as one native Rust program in which every action is a command. Each project borrows ideas from the other.

Which one should I start with? If you want to open a dataset in a browser tab or on a phone right now, or work in Jupyter or R, start with GeoLibre. If you want a desktop GIS with a ribbon, a command line that can do everything the interface can, or reports and data design, start with RustGIS.

Do they run the same geoprocessing tools? Yes: the GeoLibre and Whitebox suite of over 1,000 tools, built from geolibre-rust. RustGIS adds its own built-in vector tools and guided workflows; GeoLibre adds its Toolbox (routing, H3 grids, spatial statistics).

Can a GeoLibre plugin run in RustGIS, or the other way round? No. GeoLibre plugins are TypeScript modules for its web interface; RustGIS plugins are sandboxed WebAssembly modules that add commands and tools.

Can I open a GeoLibre project in RustGIS? Not yet. Export the layers (GeoPackage or GeoParquet keep the most) and add them to RustGIS.

See also¶

• Full comparison: RustGIS beside GeoLibre, QGIS, ArcGIS Pro and ArcGIS Online
• RustGIS and QGIS: a closer look, for QGIS users
• GeoLibre and its comparison with QGIS, ArcGIS, CARTO, Felt and kepler.gl
• Analysis: the GeoLibre and Whitebox tools in RustGIS