Flai in QGIS: open data and AI analysis one click away

marketing96143
Sep 10
4 min read

Updated: Sep 17

QGIS is a leading open-source desktop GIS for Windows, macOS and Linux, used daily for cartography, spatial analysis, and 2D/3D data editing and publishing [1]. With its core features and an ever-expanding plugin ecosystem - now 2,500+ community add-ons - QGIS covers everything from data viewing and editing to advanced map production [1, 2].

QGIS’s Processing framework integrates libraries like GDAL, GRASS, and SAGA for GUI-driven geoprocessing - buffers, clips, terrain tools, and more [3]. It likewise supports point-cloud data end-to-end: load LAS/LAZ files and it builds an EPT index for fast 3D rendering, styling, and analysis [4]. QGIS/PDAL also support Cloud-Optimized Point Clouds (COPC) - a single-file LAZ 1.4 organized as a clustered octree for efficient cloud/HTTP access [5].

Building on this foundation, the Flai plugin extends the QGIS ecosystem with:

direct access to open data via DataHub, and
a GUI to run our in-house advanced geoinformatics workflows.

DataHub integration

It is a curated catalogue of publicly available, annotated point-cloud datasets from regions worldwide and across multiple acquisition periods. It’s available on our hub.flai.ai, but we also brought it into QGIS so you can discover, select, download, and load datasets into your map in just a few clicks.

As shown in this video, you can for example select the region on a map and trigger the automatic download of the open LIDAR data directly from the plugin for Slovenia, USA, Germany, UK, Poland, Finland or any other country.

Can’t find your favorite public dataset? Let us know and we’ll add it to the catalogue!

Flai self-hosted solution integration

Behind the scenes, the plugin talks to Flai Self-hosted solution also known as Flai CLI. For users who prefer to stay in the GUI, we provide a full QGIS interface to the CLI - no terminal required - while preserving all existing CLI capabilities.

From common workflows like hillshade and vectorization to advanced AI-based classifiers for LiDAR point clouds (semantic segmentation, object detection), you can run the same pipelines directly in QGIS. Select inputs from file browser, monitor progress and logs inside the panel, and have both inputs and outputs automatically added to your QGIS project.

Unlike DataHub, the CLI features are part of our paid offering. Don’t worry - we offer a free trial.

Start free trial

Don’t forget to checkout out our official docs to see if your system has everything required to run our CLI (like wsl).

Plugin know how

Getting started

Install QGIS. Download it from the official QGIS website, install via your OS package manager (e.g., apt, Flatpak), or use Conda/Mamba (recommended on Linux - check tutorial here).
Pick a version. We recommend the LTR (Long Term Release) for stability.
Install the Flai plugin. Open QGIS > Plugins > Manage and Install Plugins... > search for Flai > select it > Install.
Initialize the plugin. Click Hard reset (the button in the plugin toolbar) to perform the initial setup, then click Show to open the main panel. The UI will guide you through the steps.
Install the Flai SDK (first launch). On first open, you’ll be prompted to install the Flai SDK. Click Install - the plugin will handle the download and configuration automatically. When the prompts finish, the plugin is ready to use.

Menu / toolbox buttons

These buttons let you show/hide the plugin panel, perform a hard reset (useful after failures or to stop a running CLI flow), and toggle the startup warning.

Our buttons can be found on the left side of the QGIS and after clicking Menu > Plugins

GUI overview

In the DataHub tab you can discover and download open datasets. The Processing Engine tab lets you run Flai-CLI workflows and view logs (it’s locked until Flai-CLI is installed or selected on your system). In Settings, you can configure/reset paths, open working folders, and update the Flai SDK.

Docking behavior

By default, the Flai plugin opens as a docked panel inside QGIS, so it’s visible whenever you bring QGIS to the foreground (e.g., by clicking the QGIS icon on your taskbar/Dock). If the UI feels cramped, you can undock the panel and move it anywhere on your screen.

When the panel is docked, double-click its title bar to snap it to the intended (recommended) size defined by the developers. The same double-click also toggles docking: double-click once to undock; double-click again to re-dock the panel to its previous region at the intended size.

Updating SDK

As our codebase grows, new capabilities will roll out to the plugin. When a feature requires a newer Flai SDK, the plugin will notify you and run the upgrade automatically.

If you suspect something is wrong with your SDK, open Settings > Update flai-sdk package to reinstall it.

You can also update the package manually, but you’ll need to know which Python environment QGIS uses and where it’s located. Step-by-step instructions are in the plugin README and the SDK README.

Summary

Flai plugin greatly contributes to open data accessibility in QGIS and enhances QGIS’s processing power with Flai CLI integration. It makes acquiring and displaying data simple and offers users complete processing workflow with just few clicks all in one software while not needing any technical knowledge of setting up environment - plugin automates everything for users.