Cloud Electronic Structure
Online DFT calculations in your browser
Execute rigorous density functional theory calculations directly in your web browser. ChemDFT couples an interactive molecular workstation with verified scientific engines running on isolated cloud workers, governed by a strict Result Firewall.
Sign in with Google. No credit card required. Free for researchers during the beta.
What you can run, and the limits
ChemDFT provides hosted cloud compute for quantum chemistry and materials modeling within strictly defined, transparent boundaries. Calculations execute in isolated Linux worker environments on Amazon Web Services (AWS) pilot hardware, automatically provisioning on demand and shutting down after approximately fifteen minutes of idle time. Because computational resources are shared and free during the beta, hosted system size limits are modest and strictly enforced before submission.
For molecular quantum chemistry using PySCF 2.14.0, the hosted size limit is 12 atoms for geometry optimisation and molecular orbital calculations, and 6 atoms for harmonic vibrational frequencies and time-dependent DFT (TDDFT). For semi-empirical calculations using xTB 6.7.1, the hosted limit is 50 atoms. For conformational search using CREST 3.0.2, the hosted limit is 12 atoms for neutral singlet hydrocarbons and oxygenates. For periodic solid-state calculations using Quantum ESPRESSO 7.6 or periodic PySCF, the unit cell limit is 20 atoms.
ChemDFT does not replace high-performance computing clusters or supercomputers for large macromolecules, heavy polymers, or broad crystal supercells. Every result is provided as a transparent research aid that researchers must independently check and verify before publication. The artificial intelligence assistant plans workflows and explains findings, but it never writes, interpolates, or modifies a numerical calculation result.
Supported molecular DFT functionals in PySCF include HF, B3LYP, PBE paired with Gaussian basis sets STO-3G, 6-31G, def2-SVP, def2-SVPD. In periodic Quantum ESPRESSO 7.6, the PBE functional and Dudarev PBE+U are supported with SSSP PBE Efficiency v2.0 pseudopotentials on k-point grids up to 4×4×4.
Understanding Density Functional Theory in computational research
Density Functional Theory (DFT) is the cornerstone of modern computational chemistry and solid-state physics. Grounded in the Hohenberg–Kohn theorems and the Kohn–Sham formulation, DFT proves that the ground-state properties of a many-electron system are completely determined by its electron density distribution rather than its multi-dimensional wavefunction. By replacing the intractable 3N-dimensional electronic Schrödinger equation with single-particle Kohn–Sham equations, DFT makes quantitative quantum mechanical modeling computationally accessible for molecules and solids.
In practical chemical investigations, DFT provides quantitative insights into ground-state total energies, equilibrium geometries, frontier molecular orbital levels, vibrational spectra, and thermochemical reaction profiles. However, standard computational chemistry workflows historically required tedious local software installation, intricate configuration of compiler flags and numerical math libraries, manual shell scripting, and command-line execution on remote clusters. ChemDFT eliminates local operational friction by bringing established open-source engines into an accessible, provenance-tracking web interface.
What ChemDFT runs: molecular and periodic engines
ChemDFT executes genuine open-source scientific software packages inside containerized cloud environments. For molecular electronic structure, ChemDFT runs PySCF 2.14.0, supporting Restricted and Unrestricted Hartree–Fock (HF), the Becke 3-parameter Lee–Yang–Parr hybrid functional (B3LYP), and the Perdew–Burke–Ernzerhof generalized gradient approximation (PBE). Integrals and exchange-correlation evaluations are powered by libcint and Libxc. Standard basis sets include STO-3G, 6-31G, def2-SVP, def2-SVPD, allowing systematic convergence from minimal representations to polarized valence double-zeta descriptions.
For rapid preliminary screening and conformer generation, ChemDFT integrates Grimme group tools: semi-empirical GFN2-xTB 6.7.1 for rapid structure optimization and CREST 3.0.2 for stochastic conformational exploration of neutral singlet molecules. For condensed-phase materials, ChemDFT hosts Quantum ESPRESSO 7.6, providing plane-wave periodic DFT using the Standard Solid State Pseudopotentials (SSSP) PBE Efficiency v2.0 library, with Hubbard U corrections for correlated transition metals, electronic density of states, and high-symmetry band structure paths.
How a calculation runs: the Scientific Result Firewall
Every job on ChemDFT follows a reproducible, auditable workflow designed around scientific integrity. First, the user prepares or imports a molecular or crystal structure. The AI Scientist can suggest appropriate methodologies or help inspect structural connectivity, but the researcher retains full control over the electronic configuration, charge, spin multiplicity, method, and basis set.
Before any compute executes, the system generates an explicit approval card detailing the normalized input hash, estimated runtime, and compute reservation. When the user approves, the job is dispatched to an isolated Linux container on an AWS EC2 instance. Upon completion, raw output artifacts pass through ChemDFT's Scientific Result Firewall. The firewall checks self-consistent field (SCF) convergence, verifies producer authority, confirms hash consistency, and permanently locks the calculation results. Numerical values are read directly from engine output files; the AI model is strictly prohibited from authoring, altering, or estimating numbers.
Who ChemDFT is designed for
ChemDFT is tailored for researchers, educators, and students who need reliable quantum chemical calculations without the overhead of maintaining local software stacks or submitting jobs through complex cluster schedulers. Synthetic organic and inorganic chemists can quickly evaluate reaction thermochemistry, compare infrared spectra, or examine HOMO–LUMO energy gaps for proposed intermediates. Materials scientists can inspect crystal unit cells and calculate band structures for small periodic systems.
ChemDFT serves as an exploratory research workstation and educational environment. Because hosted size limits are strictly capped, researchers handling large supramolecular assemblies, solvated biopolymers, or complex defect supercells should continue utilizing specialized institutional high-performance computing clusters.
Frequently Asked Questions
Is ChemDFT really free to use during the beta?
Yes. Scientists and students can sign up using a Google account and run calculations at no financial charge during the beta period. There are no credit cards required, and per-user daily limits (10 compute jobs and 40 AI runs per day) ensure fair access.
Can I run calculations on large proteins or polymers?
No. ChemDFT enforces strict hosted limits: PySCF calculations are limited to 12 atoms for optimizations/orbitals and 6 atoms for frequencies/TDDFT, while xTB supports up to 50 atoms. Large macromolecules exceed hosted capacity and should be computed on local HPC clusters.
Does the AI assistant calculate or fabricate the numbers?
Never. The AI Scientist plans workflows and explains terminology, but it is architecturally separated by the Scientific Result Firewall. Every number displayed comes strictly from deterministic open-source engine output files.
How do I cite calculations run on ChemDFT?
Each calculation provides complete provenance, including the exact software version, pseudopotentials, basis sets, input hashes, and primary literature DOIs. Visit our How to Cite guide for standard BibTeX and publication formats.