AFLOW

The Automatic-Flow (AFLOW) framework is a class of interconnected software tools designed for high-throughput materials discovery, leveraging both first-principles quantum mechanical calculations using density functional theory (DFT), and data informatics as a screening strategy.

In tandem, aflow.org, a web ecosystem of applications mirroring the functionality of AFLOW, and databases following the FAIR (Findable, Accessible, Interoperable, and Reusable) principles provide both graphical and programmatic methods to access over 3.5 million material entries. The databases also contain over 2000 crystallographic prototypes, used to uniquely classify crystal structures. For users who require only a subset of AFLOW features, such as to create convex hulls for thermodynamic analysis, or who are unfamiliar with command line interfaces, aflow.org is an ideal choice to find the necessary information for their research.

---

Quick Start

The easiest way to install aflow is to utilize our binary releases via the different package managers or as direct download.

Linux

snap

Snap packages are a modern package format designed for Linux, developed by Canonical. They encapsulate software and all its dependencies, making it easy to install applications without worrying about system-wide conflicts or missing libraries. It ensures also that aflow is kept up to date. Snaps are secure by default, running in isolated environments (sandboxes), and they work across various Linux distributions. The drawback is that aflow can only interact with files in your home directory.

snap install aflow

.deb (Ubuntu & Debian)

.deb packages are the standard software package format for Debian-based Linux distributions. They contain precompiled binaries. We pre-build .deb packages for different systems, and they can be downloaded from the GitHub Release page.

They can either be installed by double-clicking on an Ubuntu desktop system or using dpkg on the commandline.

curl -O https://github.com/aflow-org/aflow/releases/download/v4.0.5/aflow-4.0.5-ubuntu24-amd64.deb
sudo dpkg -i ./aflow-4.0.5-ubuntu24-amd64.deb

.sh (Self-extracting installer script)

This self-extracting installer allows for an interactive installation experience by simply running the following:

curl -O https://github.com/aflow-org/aflow/releases/download/v4.0.5/aflow-4.0.5-ubuntu24-amd64.sh
sh aflow-4.0.5-ubuntu24-amd64.sh

Use the --help option to see available installation options.

MacOS

Homebrew is a widely-used and reliable package manager for macOS that simplifies installing software. It allows you to easily install tools and applications without manually downloading or compiling them.

brew tap aflow-org/aflow
brew install aflow

Optional dependencies

To use all export functionality of aflow some optional dependencies need to be installed. These are gnuplot, TeXLive, and imagemagick.

On Debian based systems they can be installed with:

sudo apt update
sudo apt install gnuplot imagemagick texlive-full

and under macOS using Homebrew.

brew install gnuplot imagemagick
brew install --cask mactex

Installing from source

aflow can also be built directly from source. This method allows users to modify the source code and is recommended for the best possible performance.

Get the source code

If git is available, it is recommended to download the source code directly from GitHub.

git clone --recurse-submodules https://github.com/aflow-org/aflow.git

Otherwise, the full source can be downloaded on the GitHub Release page.

curl -O https://github.com/aflow-org/aflow/archive/refs/tags/aflow_4.0.5.orig.tar.xz
tar -xzmf aflow_4.0.5.orig.tar.xz

Build tools

To build aflow you need cmake, ninja, and a C/C++ compiler.

On Debian based systems they can be installed with:

sudo apt update
sudo apt install build-essential cmake ninja-build pkg-config

In case the available versions of the build tools provided by your operating system are too old, install the tools locally from the cmake or ninja release pages.

On macOS, the C/C++ compiler is provided by the Xcode Command Line Tools, which are installed with Homebrew. Homebrew is used to install the other build dependencies.

brew update
brew install cmake ninja

Dependencies

aflow utilizes a few external libraries that can either be provided system-wide as shared libraries or built alongside aflow using the vcpkg package manager. Using system-wide libraries reduces the build times and potentially reduces the memory footprint of aflow instances. Utilizing vcpkg will reduce the dependency on the host system which is especially helpful on HPC systems.

The following libraries are used:

• libarchive
  • for compression like xz and tar
• OpenSSL
  • for hashing function
  • for SSL & TLS connections
• libcurl
  • for querying AFLOW API

To install dependencies system-wide on Debian based systems:

sudo apt update
sudo apt install libarchive libssl-dev libcurl4-openssl-dev

To install dependencies system-wide on macOS:

brew update
brew install libarchive openssl@3 curl

For the static approach vcpkg needs to be available. It can be set up locally. It is important that vcpkg is initialized and the VCPKG_ROOT environment variable is set.

Building and installing aflow

First the buildsystem needs to prepare the system. Different presets are available that already cover most needed cases.

• all available presets can be listed with: cmake -S src --list-presets
• presets including release are optimized builds intended for production use, while debug should be used during development
• presets with vcpkg will build a static linked version of aflow that can be helpful for computer clusters, or where libraries can't be easy installed or requested by users
  • for all vcpkg presets the environment variable VCPKG_ROOT needs to be set
• customized presets can be added to a CMakeUserPresets.json (see CMakeUserPresets.example.json for an example)

To generate aflow using shared libraries follow these steps:

• create and set up the build folder with:
  • cmake -S src --preset release
  • this step configures the build environment with the available tools (compilers, linkers ...)
  • it will inform you if there are missing dependencies
• build the aflow executable in parallel with 8 threads:
  • cmake --build build/release --parallel 8
• aflow can then be executed:
  • build/release/aflow --version
  • ctest --test-dir build/release/ -L quick --parallel 8
• if the binary works as expected, it can be installed with:
  • sudo cmake --install build/release/

To generate aflow alongside the needed libraries use the release_vcpkg preset:

cmake -S src --preset release_vcpkg
cmake --build build/release_vcpkg --parallel 8
build/release_vcpkg/aflow --version
ctest --test-dir build/release_vcpkg/ -L quick --parallel 8
sudo cmake --install build/release_vcpkg/

Troubleshooting

• if the vcpkg build fails due to problems with an external library:
  • remove the binary cache of vcpkg: rm -rf ~/.cache/vcpkg/archives/
  • remove the build folder: rm -rf build/release_vcpkg
  • rerun the cmake preparation step: cmake --preset release_vcpkg
  • note: this problem can happen if the compiler or base libraries on an HPC machine changes
• Make sure you have no spaces in the path to vcpkg or the project. Vcpkg will complain and not work.

Testing

Multiple tests are defined in CMakeLists.txt to check the integrity of the build aflow binary. When aflow is built from source the full test suite can be run in the build directory. For a quick check a special test group is available: ctest -L quick --parallel 8, while ctest -E 'ut.structure_gen.proto|ut.aflowlib.lib2raw' --parallel 8 should be used for a detailed test run. If a test failed, ctest --rerun-failed --parallel 8 --extra-verbose reruns only the failed test with more details.

If aflow is installed from a binary release, a subset of tests can be run by using: aflow --unit_test. A common quick test would be: aflow --unit_test aurostd.

Documentation

The usage of aflow is described in multiple peer-reviewed publications. These should also be cited if aflow is used in any scientific context.

AFLOW4: Heading Toward Disorder
S. Divilov, H. Eckert, S.D. Thiel, S.D. Griesemer, R. Friedrich, N.H. Anderson, M.J. Mehl, D. Hicks, M. Esters, N. Hotz, X. Campilongo, A. Calzolari, and S. Curtarolo
High Entropy Alloys & Materials (2025)
DOI: 10.1007/s44210-025-00058-2 | BibTeX

aflow++: A C++ framework for autonomous materials design
C. Oses, M. Esters, D. Hicks, S. Divilov, H. Eckert, R. Friedrich, M.J. Mehl, A. Smolyanyuk, X. Campilongo, A. van de Walle, J. Schroers, A.G. Kusne, I. Takeuchi, E. Zurek, M.B. Nardelli, M. Fornari, Y. Lederer, O. Levy, C. Toher, and S. Curtarolo
Computational Materials Science 217, 111889 (2023).
DOI: 10.1016/j.commatsci.2022.111889 | BibTeX

The AFLOW standard for high-throughput materials science calculations
C.E. Calderon, J.J. Plata, C. Toher, C. Oses, O. Levy, M. Fornari, A. Natan, M. J. Mehl, G. Hart, M.B. Nardelli, and S. Curtarolo
Computational Materials Science 108A, 233-238 (2015).
DOI: 10.1016/j.commatsci.2015.07.019 | BibTeX

To document the inner working of aflow doxygen is used. Details on how to build the code documentation and style are described in docs/README.md. An online version of this documentation is available at aflow.org/aflow-documentation/.