Gaussian approximation potential (GAP) [1] for platinum. It has been fitted with QUIP/GAP [1,2] by generating a new database of atomic structures containing: dimers; fcc, bcc, hcp and simple-cubic supercells, including strained, distorted and high-temperature configurations; surface slabs; clusters (nanoparticles). Version 2 of this potential included more clusters in the training set than version 1. The calculations were carried out at the PBE level of theory [3] using the VASP code [4,5]. This potential uses 2-body (distance_2b) and SOAP-type descriptors (soap_turbo) [6,7]. The files can be used both with QUIP/GAP (compiled with the soap_turbo libraries) and TurboGAP [8]. This is the reference publication for this potential: J. Kloppenburg, L. B. Pártay, H. Jónsson, M. A. Caro. "A general-purpose machine learning Pt interatomic potential for an accurate description of bulk, surfaces and nanoparticles". J. Chem. Phys. 158, 134704 (2023) References A.P. Bartók, M.C. Payne, R. Kondor, and G. Csányi. Phys. Rev. Lett. 104, 136403 (2010). LibAtoms: https://libatoms.github.io J.P. Perdew, K. Burke and M. Ernzerhof. Phys. Rev. Lett. 77, 3865 (1996). VASP: http://vasp.at G. Kresse and J. Furthmüller. Phys. Rev. B 54, 11169 (1996). A.P. Bartók, R. Kondor, and G. Csányi. Phys. Rev. B 87, 184115 (2013). M.A. Caro. Phys. Rev. B 100, 024112 (2019). TurboGAP: http://turbogap.fi