Chemistry at HARvard Macromolecular Mechanics

Welcome to the CHARMM Development Project

CHARMM Registration

CHARMM is a versatile molecular modeling and simulation program designed to provide broad application to the simulation of many-particle systems, and includes a comprehensive set of force fields to simulate biomolecules, such as proteins, nucleic acids, carbohydrates, lipids, their assemblies, and the small molecules that interact with these targets. CHARMM runs on a variety of UNIX-compatible platforms, with optional graphical output. For more information on the CHARMM program, its capabilities and documentation, visit

We distribute the CHARMM program for academic, government and nonprofit use only. (For-profit companies may contact BIOVIA, which distributes the commercial version, called CHARMm.) A reduced version of CHARMM without high-performance parallel MD functionality is available to academic, government and nonprofit labs free of charge as charmm.

For a $600 licensing fee, individual research groups can obtain the enhanced version of CHARMM, designed to run on high-performance parallel computer platforms with integrated GPUs, as well as on smaller GPU-based workstations (using OpenMM GPU acceleration). The enhanced version is also available to computing centers under a site license for $2500, provided that only the program executable will be provided to its users, and that access to the executable will be limited to academic or nonprofit users only; users need not have individual CHARMM licenses.

CHARMM includes all of the algorithms and methodologies of charmm, with a subset of these approaches optimized for HPC/GPU computing. CHARMM and charmm use the same CHARMM scripting language, with minimal changes required to activate CHARMM HPC features. The high-performance CHARMM version requires FFTW (or Intel mkl) libraries, a recent version of MPI libraries, and one of the following supported compilers: GCC 4.8.3, Intel 14, PGI 2013 (or newer versions).

Both charmm and CHARMM can utilize OpenMM GPU acceleration for MD simulations, which require NVIDIA’s CUDA 6.5 or newer libraries, NVIDIA GPU drivers, and OpenMM version 6.2 or greater.

To register for the high-performance individual or computing center version of CHARMM (v. 46), click here.

To download the freely-available reduced version (charmm) click here.