Skip to content

xfea

The finite element model your AI writes: one Python description of models and results, the same for any solver.

Your AI assistant builds the model through xfea’s MCP server, checks it, solves it and reads the results back, with evidence an engineer can defend. The same Python is yours to read and to write.

  • One description of a model. Parts, materials, sections, supports, loads and steps, written in Python with no solver syntax in it.
  • One way to read results. Displacements, reactions and stresses come back as the same objects, whatever solver produced them.
  • Checks before and after the solve. model.check() lists what a solver would get wrong without saying so, and every solve leaves a run record and a one-page report.
  • Your solver, on your machine. The mesh and the results never leave it. Only the model’s definition goes to xfea’s conversion service, which returns the solver’s input.

The solver is yours and runs on your machine. xfea’s conversion service writes its input, with an account on app.xfea.ai: What’s free, what needs an account says what you need.

  • Core concepts: how xfea describes a structure, solves it and lets you check it.
  • Architecture: what is open, what is a service, and where each part runs.
  • Verification: xfea’s results against known answers, case by case.
  • Examples: every example script, with a picture of its model and its result.
  • API reference: generated from the code.

© 2026 Francesco Ranaudo · Text and pictures CC BY 4.0 · Code Elastic License 2.0 · how to reuse them