GeoAI

GeoAI is a free, open-source Python package that brings artificial intelligence to geospatial data. It gives researchers and GIS practitioners high-level tools to apply deep learning to satellite imagery, aerial photos and vector data — extracting building footprints, classifying land cover, detecting change — often in just a few lines of code.

Overview

GeoAI is an open-source project on GitHub with about 3,400 stars, under very active development in 2026 by the OpenGeos community. Published in the Journal of Open Source Software, the package unifies the geospatial AI workflow: searching and downloading imagery from providers like Sentinel and Landsat, preparing training datasets with image chips and labels, training segmentation, classification and detection models on top of PyTorch, and visualising results interactively with Leafmap and MapLibre. A dedicated QGIS plugin lets users run the same AI workflows from the QGIS desktop without writing code. It supports common formats including GeoTIFF, GeoJSON, Shapefile and GeoPackage, with automatic GPU device management.

Key Features

• End-to-end geospatial AI: data download, preparation, training and inference
• Segmentation workflows for buildings, water, wetlands, solar panels and more
• Pre-trained models for land cover and land use classification
• Change detection and object detection in aerial and satellite imagery
• Interactive visualisation via Leafmap and MapLibre
• QGIS plugin for no-code geospatial AI inside the desktop GIS

💳 Pricing

GeoAI is completely free and open source under the MIT license. Install it with pip or conda and run it on your own machine — there are no fees or paid tiers.

Who Is It For

GeoAI is for geospatial researchers, remote-sensing analysts, urban planners and environmental scientists who want modern deep-learning methods for Earth observation data without stitching together a fragmented toolchain.