Latest release · v1.921 · BSD-3

Download KherveFitting

Free and open. Builds are distributed via GitHub Releases. To keep the software free of charge, no signing certificate or security key was purchased — installers and scripts are delivered unsigned.

Windows

Installer or binary

macOS

Intel / Apple Silicon

Linux

Run from source

Latest Windows installer

The Windows installer for the latest release. macOS and Linux users, see below.

Download latest Browse all releases

Windows

Option 1 — Installer

  1. Run the KherveFitting installer.
  2. Approve the unsigned executable when Windows SmartScreen prompts (More info → Run anyway).
  3. Install to the default location — Program Files is fine. From v1.90, your preferences, element library and backups are kept in your user folder, so the app works correctly there.
  4. Do not run KherveFitting as administrator — drag-and-drop file loading will not work.

Note: v1.8 and earlier had to be installed outside Program Files — from v1.90 that is no longer necessary.

Option 2 — Portable zip

Prefer no installer? Download the portable zip from the latest release, extract it, and run the executable directly — no installation required.

macOS

Mount the KherveFitting bundle, drag it to /Applications, then right-click → Open the first time so macOS Gatekeeper allows the unsigned app to run.

Linux / from source

git clone https://github.com/KherveFitting/KherveFitting.git
cd KherveFitting
pip install -r requirements.txt
python Main.py

System requirements

  • Windows 10/11, macOS 12+, or Linux (Python 3.10+ for source)
  • 4 GB RAM minimum, 8 GB recommended for large datasets / depth profiles
  • Display: 1280×720 or higher

User manual (PDF)

Full reference: peak models, backgrounds, RSF tables, quantification, and more.

Publication

Kerhervé et al., Surface and Interface Analysis, Wiley (2026) — DOI: 10.1002/sia.70032

Unsigned executable warning

KherveFitting is delivered unsigned to keep it free. Your OS may show a SmartScreen / Gatekeeper warning the first time you run it — this is expected.