The ultimate python-based modeling platform for Engineers

A general-purpose model generates a formula that looks correct. It may transpose an exponent, apply a correlation outside its valid range, or invent a coefficient — and it will do so fluently. Every model here was curated from established engineering literature, implemented to a consistent pattern, and validated against known solutions. You get an auditable building block with its source and assumptions attached, not a plausible-looking answer — the example above is one, running.

The worked examples on this site span seven disciplines — pharmaceutical, chemical and process, energy, materials, aerospace, medical and biological, and geotechnical. The library behind them is broader than that: 20,000+ validated models across engineering. It is also why search answers what a model does rather than the field someone filed it under.

Yes, and that is the only way to search a library this size. Describe the physics and the assistant returns candidates by what they do, each one a validated Python implementation with its inputs, units and source already declared. Pick one and it lands on the canvas as a node, ready to wire — how it works follows that from the search box to a running model.

That is what the canvas is for. Models are nodes, outputs wire into inputs, and the whole graph runs as one calculation — so a multi-step process ends up as a single thing you can run, share and re-run against new numbers, rather than a chain of spreadsheets nobody else can open. The explore pages are exactly that, running live: drug dissolution, pan coater, solar battery.

Neither. The canvas finds, wires and runs models without code, and nothing is installed — it all happens in the browser. Being able to read Python helps: when you want to look underneath, or change something, the code is right there rather than hidden behind a licence. It is real Python rather than a dialect, so the same file runs on your own machine. The canvas on how-it-works is the live one, with no account and nothing to install first.

Yes, and you keep them. Port existing Python into a calculator and it becomes searchable, versioned and runnable alongside everything in the library, with its inputs and units declared once instead of remembered. It does not stop being yours.

No — there is no proprietary runtime to be locked into. A calculator is standard Python, and what you export is the same code that ran here. That is deliberate: a model you cannot take with you is a model you should not build your engineering on.

Point the model generator at a paper or a PDF and it will produce a Python implementation you can inspect, correct and validate before use — or write the calculator yourself and keep it private to your account. If what you need is bigger than one model, the team that wrote LIGGGHTS® and Aspherix® builds them for a living.

On servers in Germany, under EU data protection law. The privacy policy is the specific answer: what is stored, for how long, and on what legal basis.

DCS Computing GmbH in Linz, Austria — the team behind LIGGGHTS®, the open-source DEM particle simulation code, its commercial successor Aspherix®, and CFDEMcoupling, the open-source CFD-DEM framework built on OpenFOAM.