Awesome GitHub RepositoriesInteratomic Potentials

This project is a parallel simulation engine and molecular dynamics simulator designed to model the physical movements of atoms and molecules. It functions as an interatomic potential framework for calculating forces between particles and a materials analysis tool for computing thermodynamic, structural, and transport properties of solids and fluids. The engine is distinguished by its high-performance computing capabilities, utilizing spatial-domain decomposition and message-passing interface communication to distribute workloads across processors. It supports multi-backend GPU acceleration v

ocp is the Open Catalyst Project's library of state-of-the-art machine learning algorithms for catalysis.

A deep learning package for many-body potential energy representation and molecular dynamics

MACE - Table of contents - About MACE - Documentation - Installation - pip installation - pip installation from source - Usage - Training - Evaluation - Tutorials - CUDA acceleration with cuEquivariance - Weights and Biases for experiment tracking - Pretrained Foundation Models - MACE-MP:…

NequIP is an open-source code for building E(3)-equivariant interatomic potentials.

TorchANI 2.0 is an open-source library that supports training, development, and research of ANI-style neural network interatomic potentials. It was originally developed and is currently maintained by the Roitberg group.

This package implements the Allegro E(3)-equivariant machine learning interatomic potential.

TorchMD-NET provides state-of-the-art neural networks potentials (NNPs) and a mechanism to train them. It offers efficient and fast implementations if several NNPs and it is integrated in GPU-accelerated molecular dynamics code like ACEMD, OpenMM and TorchMD. TorchMD-NET exposes its NNPs as…

The Neural Force Field (NFF) code is an API based on SchNet 1-4, DimeNet 5, PaiNN 6-7 and DANN 8. It provides an interface to train and evaluate neural networks for force fields. It can also be used as a property predictor that uses both 3D geometries and 2D graph information 9.

n2p2 - A neural network potential package

For more details visit: sgdml.org Documentation can be found here: docs.sgdml.org

autoplex is still under very active development and larger modifications to the source code should be expected.

A python library for calculating materials properties from the PES

PiNN 1 is a Python library built on top of TensorFlow for building atomic neural network potentials. The PiNN library also provides elemental layers and abstractions to implement various atomic neural networks.

graph-pes is a framework built to accelerate the development of machine-learned potential energy surface (PES) models that act on graph representations of atomic structures.

📘Documentation | 🛠️Installation | 📜Recipes | 🚀Quickstart

The goal of this project is to promote the use of neural network potentials (NNPs) by providing highly optimized, open source implementations of bottleneck operations that appear in popular potentials. These are the core design principles.

Update 06/2026: Faster Molecular Dynamics with Neural Network Potentials via Distilled Multiple Time-Stepping and Nonconservative Forces Check the DMTS-NC paper (J. Chem. Theory Comput., 2026, DOI: 10.1021/acs.jctc.6c00653)). DMTS-NC delivers a 15% 𝐭𝐨 30% 𝐬𝐩𝐞𝐞𝐝𝐮𝐩 𝐨𝐯𝐞𝐫…

Train, fine-tune, and manipulate machine learning models for atomistic systems

PyNEP is a python interface of the machine learning potential NEP used in GPUMD.