Hi all:
Back in January, I announced the personal project to host Gazebo Open Container Images – OCI (Docker-compatible) images for all active Gazebo releases. Today we’re happy to share that these images have moved to the official openrobotics GitHub organization and are now adpoted and maintained by the OSRA Infrastructure team 
This move marks an important maturity milestone for the project: the Gazebo container images are no longer an individual effort but an officially supported piece of the Gazebo ecosystem infrastructure, with the same level of commitment you expect from other OSRA-maintained projects.
What’s New
- Official home: The repository now lives at github.com/openrobotics/gz_oci_images
- Official registry: All images are published under
ghcr.io/openrobotics/gazebo
Migration from ghrc.io/j-rivero images
The container images are being hosted per GitHub organization / per GitHub personal account, unfortunately there is no automatic migration path that can be established for redirecting images from my personal account registry to the openrobotics registry so please update the URIs to point to ghcr.io/openrobotics/gazebo.
What’s Available
The images cover all current Gazebo releases:
- Jetty (LTS)
- Ionic
- Harmonic (LTS)
- Fortress (LTS)
Each release is available in two variants:
core- Minimal installation with essential tools (up to sdformat)full- Complete Gazebo suite with all packages
All images support both linux/amd64 and linux/arm64 architectures. They are automatically rebuilt weekly and whenever new package versions are detected in the Gazebo repositories.
Quick Start with Rocker
Here’s an example using rocker to run Gazebo with display and GPU support:
# Install rocker (requires Docker)
python3 -m venv venv && source venv/bin/activate && pip install rocker
# Run Gazebo with GPU and X11 display support
rocker --x11 --nvidia gpus ghcr.io/openrobotics/gazebo:jetty-full -- gz sim --verbose
Troubleshooting
Having issues with graphics or GPU acceleration? See the Troubleshooting Guide for step-by-step diagnostics.
Attribution
This project is a fork of the work done by Shane Loretz (@sloretz) on ROS OCI Images, adapted for Gazebo. All credit for the original approach goes to Shane.