Download and install

Build from source (Linux/macOS/Windows)

The source is of Cadabra is distributed via github, see the kpeeters/cadabra2 repository for more details, including the pre-requisites.

You can build .deb and .rpm packages yourself when building from source, by running sudo cpack after a successful build.

Binary packages

Packages for various Linux distributions are available for download from github. These are generated automatically on every release. Installation on macOS can be done through HomeBrew or MacPorts; see below. There is also a Docker image if you only want to use Cadabra via Jupyter.

Use these packages at your own risk. Because of time constraints some of these packages receive almost no testing, so any feedback is welcome. If you want to help getting these into official distribution repositories, please let me know.

The AppImage version of Cadabra can be installed on any reasonably modern Linux distribution and it is completely self-contained. Just download and make the file executable with chmod u+x *.AppImage. Versions are provided for x86_64 and aarch64 architectures from the github release page.

Packages for Ubuntu 22.04 Jammy and 24.04 Noble are available from the github release page. Cadabra is now also available from the official repositories, by installing the cadabra2 package (though this may not always be the most recent version). Thanks to Alex Myczko for making this happen.

Packages for Fedora 40 are available from the github release page.

A package for openSUSE Tumbleweed is available from the github release page. You can still build it on openSUSE Leap, but you will need to build from source or rely on 3rd party repositories.

Cadabra is now available in Debian 11 (Bullseye) from the official repositories, by installing the cadabra2 package (though this may not always be the most recent version). Thanks to Alex Myczko for making this happen.

If you only want to use Cadabra from within Jupyter, you can use the official Docker image at DockerHub. This container exposes the Jupyter server on a local URL which you can access using your browser, and stores notebooks in your home directory. Available for both AMD64 and ARM64 architectures. Be aware that you will miss out on some of the goodies of the Cadabra notebook interface.

macOS

Cadabra is available via the Homebrew package manager. Simply do brew tap kpeeters/repo brew install cadabra2 If you encounter problems, please email info@cadabra.science so they can be fixed and other people can benefit from it.

Cadabra is also available via MacPorts, sudo port install cadabra2 should do the trick. For more details follow the link above.

If you want to build from source, see macOS build instructions for details.

Windows 11

A Windows installer for Cadabra is finally available again. There are some rough edges that still need to be fine-tuned, but it seems to be mostly functional, including plotting. Download the Windows installer (x86_64) from the github release page. All batteries are included so you do not need to install anything else. Many thanks to Dominic Price for doing most of the work a few years ago.

Once installed, navigate to your installation folder and start cadabra2-gtk (sorry, no start menu entry yet). There are some sample notebooks in the Examples folder.

The Conda packages are outdated and contain known bugs; if you can, avoid it and use one of the other available packages.

On Linux and macOS, you can install Cadabra, its native notebook interface, and the Jupyter kernel via Conda or Anaconda, thanks to the work of Isuru Fernando. Activate the conda-forge channel if you have not done that already, with conda config --add channels conda-forge conda config --set channel_priority strict and then install these packages by using one of the following lines: conda install cadabra2 conda install cadabra2-gtk conda install cadabra2-jupyter-kernel

Is your distribution not listed above? Either build Cadabra from source (see below) or send an email to info@cadabra.science to request a package for your distribution.