A QGIS plugin that integrates AI capabilities for geographic information system workflows and spatial data analysis.
QGIS Plugin - GeoAI is a free, open-source GIS/Mapping extension that brings deep learning and computer vision directly into QGIS workflows for geospatial analysis, with pricing starting at free. It targets remote sensing analysts, GIS professionals, environmental scientists, and researchers who need production-grade AI segmentation and detection without leaving their desktop GIS environment.
The plugin bundles multiple specialized AI panels into the QGIS interface, including Tree Segmentation powered by DeepForest, Water Segmentation via OmniWaterMask, the Moondream Vision-Language Model for natural-language image queries, Segment Anything (SamGeo) for prompt-driven masking, a combined training-and-inference Semantic Segmentation panel, and an Instance Segmentation panel built on Mask R-CNN. It supports a broad library of model architectures and encoders for segmentation tasks, multiple SAM model variants (including pending access to SAM 3), and includes a built-in dependency installer alongside Pixi-based environment setup with PyTorch and CUDA acceleration. GPU memory management tools and a plugin update checker are bundled to keep long inference sessions stable.
Use cases span tree canopy mapping, water body and wetland detection, building footprint extraction (USA, Africa, China), solar panel and parking-spot detection, ship and car detection, change detection, super-resolution, and agentic workflows via STAC and catalog search agents. Compared to the other GIS/Mapping tools in our directory, QGIS Plugin - GeoAI stands out as one of the few completely free, locally executed AI integrations for QGIS, supporting more than 60 documented example workflows and over 40 API modules. Based on our analysis of 870+ AI tools, this plugin is unusually deep for an open-source offering, rivalling commercial ArcGIS deep-learning toolboxes while running entirely on the user's machine with zero subscription cost.
Was this helpful?
Wraps the DeepForest deep-learning library to detect individual trees in aerial and drone imagery directly inside QGIS. Multiple pretrained models are supported, allowing users to run inference on RGB orthomosaics with no manual labeling. Output is delivered as standard QGIS vector layers, ready for forestry, biodiversity, and carbon-stock analysis.
Brings the Segment Anything Model family â including support for SAM 3 â into QGIS for prompt-driven segmentation of any object visible in raster imagery. Users can click points, draw boxes, or use text prompts to generate accurate masks without training a custom model. This dramatically speeds up digitization workflows that previously required hours of manual editing.
Provides a single panel that lets users create training data, train a segmentation model, and run inference without leaving QGIS. It supports multiple model architectures and encoders from the segmentation_models.pytorch ecosystem, plus modern backbones like DINOv3, Prithvi, and Tessera. This makes it practical for analysts to build custom land-cover or thematic models on their own labeled data.
Integrates the Moondream VLM so users can ask natural-language questions about map tiles and aerial imagery directly in QGIS. It supports image captioning, sliding-window analysis for large rasters, and a GUI for iterative prompting. This bridges traditional GIS analysis with modern multimodal AI for tasks like rapid scene description and content-based search.
Includes a Clear GPU Memory utility to recover VRAM between heavy inference jobs, preventing common out-of-memory crashes during long sessions. A built-in dependency installer plus a documented Pixi-based environment workflow handle the typically painful PyTorch + CUDA + DeepForest install path. Together they make QGIS-based deep learning materially more reliable than ad-hoc setups.
Free
Ready to get started with QGIS Plugin - GeoAI?
View Pricing Options âWe believe in transparent reviews. Here's what QGIS Plugin - GeoAI doesn't handle well:
Weekly insights on the latest AI tools, features, and trends delivered to your inbox.
Recent additions documented on the site include support for SAM 3 (with access request), expanded model integrations such as DINOv3 (visualization, wetlands, fine-tuned segmentation), Prithvi, Tessera, AlphaEarth, RF-DETR detection, Google Satellite Embedding, and Lightly self-supervised training, plus AI agents for STAC and catalog search, image captioning, super-resolution, field boundary detection, and a 2025 GeoAI Workshop series (AWS 2025, TNView 2025, CANVAS 2025, AGU 2025).
No reviews yet. Be the first to share your experience!
Get started with QGIS Plugin - GeoAI and see if it's the right fit for your needs.
Get Started âTake our 60-second quiz to get personalized tool recommendations
Find Your Perfect AI Stack âExplore 20 ready-to-deploy AI agent templates for sales, support, dev, research, and operations.
Browse Agent Templates â