Import legacy models
If you have a model from an earlier version of ProcessModel, you can bring it into the new version and pick up where you left off. The import reads your existing model, rebuilds it on the canvas, and shows a summary of exactly what came across before anything loads.
How to import
Section titled “How to import”Before you start, save and close any model you have open. The import replaces whatever is on the canvas, so it runs on an empty canvas.
- Open the File menu and choose Import Legacy.
- Pick your model file. Models from earlier versions of ProcessModel use the .spg extension.
- Review the import summary. It shows everything that was found and flags anything that needs a second look.
- Open the model to load it onto the canvas.
The Build with AI window also accepts a .spg file, if that is where you are starting.
What to check after importing
Section titled “What to check after importing”Before the model loads, ProcessModel shows an import summary so you know exactly what you are getting. It includes:
- Counts of everything found: steps, buffers, resources, items, routes, arrivals, schedules, and scenarios.
- What was recovered from your original model.
- What was adapted to work in the new version.
- Anything missing, for example a shift file the model refers to but that was not included in the package.
- The busiest resource, so you can sanity-check the model at a glance.
A few things are brought across for you to confirm. The summary flags each one:
- Else routes arrive as a percentage route and may need a condition added.
- Batching is carried across as a note rather than rebuilt, so you can recreate it the way you want.
- Subcharts are not brought in yet.
- Ordered arrival buffer sizes may need raising to hold the volume you expect.
- Activity and entity costs come across as 0, so re-enter them.
A few parts of an older model are not carried across. The summary flags these too:
- Action logic on a continuous arrival.
- Report statements.
- A standalone WAIT.
- OwnedResource() and Animate().
What converts
Section titled “What converts”Most of your model comes across cleanly and is ready to run. This includes:
- Entities and activities, with their positions on the canvas preserved.
- Storages, which become buffers and queues.
- Resources, including hourly cost.
- All routing: percentage, conditional, create, attach, detach, renege, and alternate.
- All arrival types: continuous, periodic, scheduled, daily pattern, and ordered (reorder).
- Attributes, including enumerated attributes, which become text.
- Variables.
- Scenario settings, resolved to their values, and any named scenarios, which are imported turned off.
- Run length.
- Shift schedules, including breaks, split shifts, and overnight shifts.
- Distributions and action logic.
Requirements and notes
Section titled “Requirements and notes”- Start on an empty canvas. The import replaces the current model, so save and close your work first.
- Older models use the .spg extension. That is the file you select when you import.
- The result is the same every time. No AI is involved in the import, so the same file always produces the same model.
- An unfamiliar file is refused, not half-imported. If ProcessModel does not recognize the format, it stops and tells you clearly instead of loading a partial model, so you never end up with an incomplete import.

