Breaking

GeoAgent: A Multimodal AI Agent for QGIS Workflows

June 29, 2026 at 08:44 EDT

AI Agents
Open Source
Software Dev & Coding

The opengeos project has released GeoAgent (its QGIS plugin edition called OpenGeoAgent), a multimodal AI agent that lets users drive QGIS and geospatial Python packages through natural language. Accepting both images and text as input, it can have the agent automatically carry out GIS workflows, with development hosted on its GitHub repository.

May–June 2026 · Open Source Release

GeoAgent: a multimodal AI agent that turns plain language into geospatial analysis

An open-source agent automates satellite imagery analysis and visualization from text and images — usable from a QGIS plugin, Jupyter, or Python scripts. Tasks like watershed analysis or Sentinel-2 mosaics run in minutes.

Environments: QGIS plugin, Jupyter, Python scripts

Supported model providers, cloud or local

Shared layer replacing per-library agents

From instruction to reproducible result

Natural language + image

→

Tool call & execution

→

Visualized result

→

Copy as Python snippet

Destructive operations prompt a confirmation dialog; any Python function becomes a tool via the @geo_tool decorator.

Integrated geospatial toolset

leafmap geemap geoai WhiteboxTools Google Earth Engine NASA OPERA / STAC QGIS

Welcomed

A genuinely open-source, multimodal agent — image attachments, screenshots, and voice input supported; can run fully local and free via Ollama.

Watch points

As a brand-new project, large-scale feedback is limited: degraded tool-calling on Ollama, accuracy on ambiguous queries, and API costs to monitor.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.