Search pages, resources, and actions
Discover models, datasets, and services for materials science.
CoolProp — an open-source thermophysical property library. Provides high-accuracy equations of state for 122+ fluids including water, refrigerants, and rocket propellants. Available in C++, Python, MATLAB, and 10+ other languages.
OpenFOAM — the leading open-source CFD toolbox. Covers incompressible/compressible flow, combustion, heat transfer, multiphase, and acoustics. Includes meshing utilities (blockMesh, snappyHexMesh) and post-processing with ParaView.
SU2 — an open-source collection of software tools for multiphysics simulation and design. Core solver handles compressible and incompressible flows, aerodynamics, and shape optimization. Widely used in aerospace engineering.
RocketPy — a Python library for simulating sounding rockets and high-power rocketry flights. Models 6-DOF trajectory, parachute deployment, motor performance, and atmospheric conditions. Includes Monte Carlo dispersion analysis.
Cantera — an open-source suite of tools for chemical kinetics, thermodynamics, and transport processes. Simulates combustion, detonations, electrochemistry, and catalysis. Supports detailed and reduced reaction mechanisms.
BOUT++ — a framework for solving differential equations on curvilinear meshes, primarily used for plasma fluid simulations in tokamak edge and scrape-off layer regions. Supports 2D and 3D simulations of plasma turbulence and transport.
PlasmaPy — an open-source Python package for plasma physics. Provides plasma parameter calculations, particle tracking, dispersion relations, and diagnostic analysis tools. The community standard Python library for plasma research.
WarpX — an exascale electromagnetic particle-in-cell (PIC) code for plasma simulations. Developed by DOE/LBNL. GPU-accelerated, supports mesh refinement, and scales to the largest supercomputers. Used for laser-plasma interactions, accelerator design, and astrophysical plasmas.
NekRS — a GPU-accelerated spectral-element CFD solver for thermal-hydraulics simulations. Excels at high-fidelity turbulent flow modeling in nuclear reactor geometries. Built on Nek5000 with OCCA GPU portability layer.
Multiphysics Object-Oriented Simulation Environment — Idaho National Laboratory's finite-element framework for coupled multiphysics problems. Powers BISON (fuel performance), Griffin (reactor physics), and other nuclear simulation codes.
OpenMC — a community-developed Monte Carlo particle transport code for neutron and photon simulations. Used for reactor physics, criticality safety, shielding, and depletion analysis. Supports CAD geometry and multi-group cross sections.
Materials Commons — a collaboration platform and data repository for the materials science community. Stores experimental data, computational results, and publications. Developed at University of Michigan.
Chemical Entities of Biological Interest — a freely available dictionary of molecular entities focused on chemical compounds. Maintained by the European Bioinformatics Institute. 60K+ entities with structures, roles, and classifications.
MatOnto — a materials science ontology for semantic data integration. Maps concepts across materials databases, enabling interoperability between different data sources and vocabularies. OWL format.
European Materials Modelling Ontology — a top-level ontology for materials science and engineering. Provides a formal framework for representing materials, models, manufacturing processes, and characterization methods. OWL format.
Stanford Synchrotron Radiation Lightsource beamline access. Book beam time for XRD, XANES, EXAFS, and other synchrotron characterization techniques. Coming soon.
Autonomous materials synthesis laboratory at UC Berkeley / LBNL. AI-driven robotic synthesis of inorganic materials. Submit target compositions for autonomous synthesis and characterization. Coming soon.
Cloud-based DFT compute cluster. Run VASP, Quantum ESPRESSO, and other DFT codes on managed GPU/CPU nodes. Pay-per-job pricing. Coming soon.
Terminal-based crystal structure viewer. Renders 3D atomic structures as ASCII art in the terminal. Supports CIF, POSCAR, and XYZ inputs. Useful for quick inspection over SSH.
Command-line tool for validating Crystallographic Information Files (CIF). Checks syntax, data names, loops, and values against IUCr CIF dictionaries. Reports errors and warnings.
MCP server for searching PubChem's database of chemical compounds, substances, and bioassays. Query by name, formula, SMILES, InChI, or structure similarity. Access 116M+ compounds.
MCP server exposing the Materials Knowledge Graph — 70K+ entities and 5.4M triples extracted from materials science literature via NLP. Query relationships between materials, properties, synthesis methods, and applications.
MCP server that federates queries across 40+ OPTIMADE-compliant materials databases. Search for structures by composition, space group, band gap, and other properties using the OPTIMADE filter language. Aggregates results from Materials Project, AFLOW, OQMD, NOMAD, and more.
matminer featurization toolkit — computes composition-based and structure-based features for ML models. Includes Magpie descriptors, orbital field matrix, Voronoi-based features, and 50+ featurizers.
Structure format converter powered by pymatgen. Converts between CIF, POSCAR, XYZ, and other crystal structure formats. Handles symmetry, oxidation states, and site properties during conversion.
Robocrystallographer — generates human-readable descriptions of crystal structures using NLP. Automatically analyzes bonding, dimensionality, and structural motifs to produce text summaries.
spglib — C library (with Python bindings) for finding and handling crystal symmetries. Determines space groups, Wyckoff positions, and symmetry operations from atomic coordinates.
Polymer Genome — a curated dataset and informatics platform for polymer property prediction. Contains 1K+ polymers with measured and predicted properties including glass transition temperature, dielectric constant, and bandgap.
The Materials Project — the canonical open database of computed materials properties. 154K+ inorganic materials with formation energies, band gaps, elastic tensors, and more. Accessible via REST API and pymatgen.
Joint Automated Repository for Various Integrated Simulations — NIST's curated database of 76K+ materials with DFT-computed properties including band structures, elastic constants, and optical properties. Includes 2D materials.
Graph Networks for Materials Exploration — Google DeepMind's dataset of 520K new stable inorganic materials discovered via active learning with GNNs. Dramatically expands known stable crystals.
Alexandria PBE dataset — 5.8 million structures with PBE-level DFT calculations from KU Leuven. Includes formation energies, band gaps, and thermodynamic stability data.
Open Materials 2024 — Meta FAIR's massive dataset of 118 million DFT-computed structures for training universal interatomic potentials. Covers diverse chemistries and structural motifs.
Matbench Discovery — an automated benchmark for ML models on materials stability prediction. Provides standardized evaluation methodology and a public leaderboard comparing UIPs and property predictors on crystal stability classification.
MODNet — Material Optimal Descriptor Network. A neural network designed for small-data materials property prediction. Uses transfer learning and composition-based descriptors to achieve strong performance with limited training data.
MatterSim by Microsoft Research — a deep-learning atomistic model across elements, temperatures, and pressures. Supports energy, force, phonon, and phase-diagram prediction.
ORB v3 by Orbital Materials — a fast, GPU-optimized universal interatomic potential for molecular dynamics. Designed for production use with emphasis on inference speed.
Atomistic Line Graph Neural Network — predicts 52+ material properties by leveraging both atom-bond and bond-angle graphs. Part of NIST JARVIS infrastructure. Top performer on the JARVIS leaderboard for multiple properties.
Crystal Graph Convolutional Neural Network — one of the earliest GNN architectures for crystal property prediction. Directly learns from crystal structure without manual feature engineering. Predicts formation energy, band gap, and stability.
MatErials Graph Network — graph neural network for predicting molecular and crystal properties. Excels at band gap and formation energy prediction. Uses global state, node, and edge features.
MACE-MP-0 — a foundation model for atomistic simulation built on the MACE equivariant message-passing architecture. Trained on MPtrj (>1M structures). 5.6M parameters. State-of-the-art accuracy on energy and force prediction across the periodic table.
Materials 3-body Graph Network — a universal potential trained on ~187K Materials Project relaxations. Supports structure relaxation, molecular dynamics, and property prediction. 225K parameters.
Crystal Hamiltonian Graph Neural Network — a universal interatomic potential (UIP) that directly learns from the Materials Project trajectory dataset. Predicts energies, forces, stresses, and magmoms in a single forward pass. 412K parameters, runs on CPU.
