A collection of tools for training and using AI models that work with geospatial and tabular data, integrating machine learning and deep learning techniques with GIS for classification, regression, and natural language processing tasks.
A collection of tools for training and using AI models that work with geospatial and tabular data, integrating machine learning and deep learning techniques with GIS for classification, regression, and natural language processing tasks.
The ArcGIS GeoAI Toolbox is Esri's integrated suite of artificial intelligence and machine learning tools built directly into ArcGIS Pro, designed specifically to bring modern AI capabilities to geospatial and tabular data analysis. Rather than requiring GIS analysts to leave their primary work environment to use external Python frameworks, the GeoAI Toolbox embeds classification, regression, clustering, deep learning, and natural language processing workflows alongside the spatial data they already manage in their ArcGIS projects. The toolbox contains over 30 geoprocessing tools organized across four toolsets — Feature and Tabular Analysis, Imagery AI, Text Analysis, and Time Series AI — and supports more than 50 deep learning model architectures including U-Net, Mask R-CNN, FasterRCNN, DeepLabV3, SSD, RetinaNet, and transformer-based models. This tight integration means trained models can be applied directly to feature classes, raster layers, and attribute tables without manual data export or format conversion.
The Feature and Tabular Analysis toolset uses automated machine learning (AutoML) to train, fine-tune, and ensemble the best-performing models on a given dataset, automating much of the model selection and hyperparameter tuning process that traditionally consumes significant analyst time. Users can perform classification and regression tasks on attribute tables and feature layers, with the toolbox handling cross-validation, feature importance reporting, and prediction generation. The Text Analysis toolset brings NLP capabilities to GIS workflows, supporting text classification, entity extraction, and sequence-to-sequence transformations, which is particularly useful when working with unstructured fields such as incident reports, social media content, or property descriptions tied to geographic locations. ArcGIS Living Atlas provides over 100 pretrained deep learning models for common remote sensing tasks, enabling users to run inference immediately without assembling training data. ArcGIS Pro is used by over 350,000 organizations worldwide across government, defense, utilities, and environmental science sectors.
Was this helpful?
Applies automated machine learning to feature classes and tables, training, tuning, and ensembling multiple models to predict categorical variables (classification) and continuous variables (regression). The automation handles algorithm selection and hyperparameter tuning so GIS analysts without ML backgrounds can still produce competitive models on their data.
Runs deep learning object detection and pixel classification on raster and imagery data, supporting workflows such as building footprint extraction, road network mapping, and land cover classification. Supports over 50 architectures including U-Net, Mask R-CNN, FasterRCNN, DeepLabV3, SSD, and RetinaNet. It leverages pretrained models from ArcGIS Living Atlas and custom models trained in the arcgis.learn module.
Performs natural language processing directly on text fields in feature classes and tables, including classification, transformation, and entity extraction for items like addresses and named places. Users can fine-tune pretrained NLP models from ArcGIS Living Atlas or train custom models using labeled text data.
Forecasts and estimates future values at specific locations inside a space-time cube, which is Esri's native structure for spatiotemporal data. This enables forecasting applications like predicting crime rates, sales, or environmental measurements at a neighborhood or sensor level rather than just in aggregate.
Models created by the GeoAI tools are fully compatible with the arcgis.learn module in the ArcGIS API for Python, so data scientists can continue training, fine-tuning, and deploying models in code. This bridges the gap between GUI-driven GIS analysts and Python-driven ML engineers working on the same project.
~$700/year per named user
~$1,400/year per named user
~$2,200/year per named user
~$1,400/year add-on
~$1,400/year each add-on
Paid (organizational quote)
Ready to get started with ArcGIS GeoAI Toolbox?
View Pricing Options →We believe in transparent reviews. Here's what ArcGIS GeoAI Toolbox doesn't handle well:
Weekly insights on the latest AI tools, features, and trends delivered to your inbox.
ArcGIS Pro 3.4 (released November 2024) expanded the GeoAI Toolbox with new transformer-based deep learning model support in the Imagery AI toolset, additional pretrained models in the Living Atlas for infrastructure and vegetation mapping, and performance improvements for GPU-accelerated training. ArcGIS Pro 3.3 (mid-2024) introduced enhanced AutoML capabilities in the Feature and Tabular Analysis toolset with support for additional ensemble methods and improved hyperparameter search. Esri's 2025–2026 roadmap includes deeper foundation model integration for geospatial AI, expanded support for large-scale distributed training, and new pretrained models targeting climate and environmental monitoring use cases. The Living Atlas now hosts over 100 pretrained deep learning models. Users should consult the ArcGIS Pro 3.4 release notes for the full list of added tools and supported architectures.
Data & Analytics
Agentic GIS Platform providing cloud-native spatial analytics that runs natively inside data warehouses like BigQuery, Snowflake, Databricks, and Redshift.
Data & Analytics
Open-source machine learning framework for developing and training neural networks and deep learning models.
No reviews yet. Be the first to share your experience!
Get started with ArcGIS GeoAI Toolbox and see if it's the right fit for your needs.
Get Started →Take our 60-second quiz to get personalized tool recommendations
Find Your Perfect AI Stack →Explore 20 ready-to-deploy AI agent templates for sales, support, dev, research, and operations.
Browse Agent Templates →