Visualize, build, and run molecular simulations — from the browser, from Python, or on your own compute infrastructure. One platform for the full molecular workflow.
Platform
Whether you prefer a visual interface, a Python script, or running on your own cluster — Atomiverse meets you where you are.
Open your browser and start working. Load structures from files or SMILES, visualize in 3D, build molecules atom-by-atom, and run MLIP-powered calculations — no installation required.
Launch the viewer →Submit calculations programmatically from any Python environment. Works with ASE Atoms objects, supports multiple levels of theory, and integrates into existing computational workflows.
View SDK docs →Install the controller on your workstation or cluster, and jobs submitted from the web app or Python SDK run on your own hardware. Use your existing resources.
Set up BYOC →Capabilities
Drag-and-drop XYZ, CIF, SDF/MOL files or paste SMILES. Rotate, zoom, and measure distances, angles, and dihedrals interactively. Customize display, overlay unit cells, and visualize non-covalent interactions.
Construct structures atom-by-atom. Add atoms or molecular fragments, adjust bond lengths, angles, and dihedrals, then optimize geometry — all inside the browser.
Single-point energies, geometry optimizations, vibrational analyses, thermochemistry, potential energy scans, and transition-state searches — powered by machine-learned interatomic potentials.
Map atoms between reactants and products, generate TS geodesic paths, analyze barrier heights, and inspect transition-state trajectories. Full reaction network generation with Autograph.
Rapidly generate conformer ensembles, filter by geometry criteria, and deduplicate structures. Essential for thorough conformational space exploration.
Scrub through multi-frame trajectories, export animated GIFs, save high-resolution PNG screenshots, and export structural data in XYZ, SMILES, and other formats.
Get started in seconds
Paste XYZ, upload a file, or type a SMILES string. In Python, pass an ASE Atoms object directly.
Inspect in 3D with measurements, or build molecules atom-by-atom with the interactive builder. Customize display, overlay unit cells, visualize interactions.
Submit MLIP-powered jobs — energies, optimizations, vibrations, conformer searches, reaction paths — from the browser or your Python script.
Run on managed compute or your own infrastructure with BYOC. Export structures, screenshots, GIFs, and results.
Python SDK
Install the SDK and submit calculations from any Python environment. Works natively with ASE Atoms objects, supports configurable levels of theory, and lets you choose where jobs run.
from ase import Atoms from atomiverse import SinglePointEnergy from atomiverse.levels import GFN2_XTB # Define a molecule using ASE atoms = Atoms("H2", positions=[ [0.0, 0.0, 0.0], [0.0, 0.0, 0.74], ]) # Submit a calculation job = SinglePointEnergy( atoms=atoms, level_of_theory=GFN2_XTB, ) result = job.submit() print(f"Energy: {result.energy:.6f} Ha")
Bring Your Own Compute
Install the lightweight controller on your workstation or HPC cluster. Jobs submitted from the web app or Python SDK are automatically routed to your infrastructure.
Run jobs on hardware you already have — workstations, lab servers, or HPC clusters. No need to provision cloud infrastructure.
One command to install, interactive setup wizard, and systemd integration. The controller runs as a background service and handles job lifecycle automatically.
Switch between managed cloud and your own compute with a single parameter. Same API, same SDK, same results — different hardware.
Resources
Guides for the web app, Python SDK, REST API, and BYOC controller setup.
Integrate parsing, conversion, job submission, and builder endpoints into your tools.
Submit jobs programmatically with ASE integration, level-of-theory configuration, and BYOC support.
Get in touch
Found a bug? Have an idea for a new feature? Or just want to say hello? We'd love to hear from you. Your feedback helps make Atomiverse better for everyone.
Open the web app or install the Python SDK — your molecules, your way.