Documentation / Quick Start

Installation

Get the Wrengler plug-in installed and running in Rhino 8 — requirements, an account, and the first launch.

ℹ️

Wrengler is in preview — there has been no release yet. The plug-in supports Rhino 8 on Windows and macOS.

Requirements

  • Rhino 8, version 8.0 or higher, on Windows or macOS.
  • A Wrengler account. The plug-in presents a sign-in wall on open and runs no study without a session — the gate applies to Rhino commands typed at the prompt as well as to the panel.
  • An internet connection at first launch, to load the panel and sign in.

The ray engine itself runs on the CPU on any machine, so every study works without special hardware.

💡

A supported GPU is a speed-up, not a prerequisite. The engine picks the best backend your machine offers and falls back on its own — there is nothing to install or configure.

Your Account

Create one on this site before you install; the plug-in's Sign up link opens your browser here anyway. Then, in the panel, sign in with e-mail and password or with GitHub — GitHub opens your real system browser and returns to Rhino on its own. An account created through GitHub has no password, so use the same button next time.

After a successful sign-in Wrengler works offline for up to 14 days before asking again. That grace window extends from a sign-in that already happened: a machine that has never signed in cannot run a study, and there is no air-gapped or trial mode.

Your model is not part of any of this — analysis runs locally on your own machine and no geometry is uploaded to be analyzed.

One sign-in covers up to 14 days offline. The panel still loads over the network each time it opens, and the model is never part of either.

Install on Windows

The route Wrengler will ship on, once released:

  1. In Rhino 8, go to Tools → Package Manager and search for Wrengler.
  2. Click Install, then restart Rhino when prompted.
  3. Type Wrengler in the command line — the panel opens, typically docked on the right, and asks you to sign in.

Updates arrive the same way: Update in Package Manager, then restart. Uninstall is Uninstall in the same place.

On Windows the engine runs on the CPU, with a BVH-accelerated backend. An NVIDIA GPU backend (OptiX/CUDA) is planned and not built yet, so an NVIDIA card does not speed studies up today.

Install on macOS

Rhino for Mac installs plug-ins as Yak packages. For a development build, use the repository's macOS installer from the project root:

scripts/build-mac-rhino.sh Release

The script builds the macOS plug-in, packages it for Rhino, and installs it with Yak. Quit Rhino before running it; then launch Rhino 8 and type Wrengler in the command line. The default backend on Apple Silicon is Metal, with CPU fallback when Metal is unavailable.

Install a development build on Windows

For a development build, until the package is published.

Yak CLI:

yak install wrengler.yak --source=.

Drag and drop (Windows):

  1. Right-click the Wrengler.rhp file → Properties → check Unblock → Apply.
  2. Drag it onto the Rhino 8 viewport and confirm.
  3. Restart Rhino.

Verify

Type Wrengler in the command line. The panel should appear, docked, showing the sign-in form (or the interface, if you are already signed in).

"An internet connection is required" — the panel is a web app the plug-in loads over the network, and it could not be reached. Check your connection or proxy and reopen the panel; nothing is wrong with the install. The engine on your machine is unaffected, but a study still needs a session, so studies stay blocked until the panel loads and you sign in.

Panel won't open, or shows blank — restart Rhino, then check the command history (_CommandHistory) for errors.

Signed in yesterday, asked again today — sessions expire; a refresh is normally invisible. If you have been offline for more than 14 days, the grace window has run out and you need a connection to sign in again.

Runs slower than expected — check the backend diagnostics to see which backend was selected, and update your GPU drivers. This only affects speed; every backend returns the same answer.

Next

First Steps — run your first study.