The structure relaxation routes now let you choose which machine-learning interatomic potential (MLIP) drives the optimization.
When you run Relax a crystal structure (or the animation / publish variants), set the model field to pick the calculator:
Orb v3 — OMat and MPA variants (conservative and direct, 20- and inf-step)
MACE — MP, MPA, and OMAT foundation models (small / medium / large where applicable)
CHGNet
If you omit model, we default to Orb v3 conservative inf MPA (orb-v3-conservative-inf-mpa), same as before.
Handy when you want to compare potentials on the same CIF, or match the model used elsewhere in a workflow. Feedback welcome if you hit edge cases with a particular structure or model combo.
Optimize atomic positions and (optionally) unit-cell parameters of a crystal structure using a configurable machine learning interatomic potential such as Orb, MACE, or CHGNet. Upload a CIF file and receive the relaxed structure as a new CIF. Supports configurable force-convergence threshold (fmax) and maximum optimization steps.
The relax routes now accept a model parameter for Orb, MACE, or CHGNet.