Search pages, resources, and actions
Discover models, datasets, and services for materials science.
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.
Materials 3-body Graph Network — a universal potential trained on ~187K Materials Project relaxations. Supports structure relaxation, molecular dynamics, and property prediction. 225K parameters.
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.
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.
