Most people who go looking for simulation software have already decided they need to buy some. That assumption is worth testing before a purchase order, because a meaningful share of the questions people buy simulation software to answer can be answered by a tool that costs nothing.
This is a comparison between the product most often described as the easiest way into discrete event simulation, and a free process modeling application that simulates further than its reputation suggests.
The short version
Bizagi Modeler is free for an individual user, and its simulation reaches resource contention and working calendars including shifts, breaks and holidays. If your question is about queueing at a handoff, staffing a service desk, or how a working week absorbs demand, that is very likely enough.
Simul8 is a discrete event tool with a deeper engine, an animated model, a browser version, and no permanent free edition at all. If your question involves variability you need to characterize properly, physical materials, batching, or a result you have to put in front of a room, that is what you are paying for.
The mistake is buying the second when you needed the first, and never finding out.
What the free one actually does
Bizagi documents its simulation in four levels, each adding information to the last.
The first validates the model: does the process reach every path, does it behave as drawn. The second adds timings and reports end to end throughput as a minimum, maximum, mean and total. The third adds resources and roles, how many there are and where they are used, which is where a model starts predicting contention rather than describing a diagram. The fourth adds calendars: shifts, day schedules, weeks, holidays and breaks.
That fourth level is the one that surprises people. A model that understands that your team is not there on Sunday, and that half of them are at lunch between twelve and one, is a materially different thing from a flowchart with numbers on it. It is also the level at which most staffing arguments are actually settled.
The published documentation carries a worked example for each level, which is unusual for a free tool and makes the ceiling easy to find for yourself.
Where that ceiling is
The four levels stop before several things a discrete event engine treats as routine.
There is no documented batching, so a process where work accumulates and is handled in groups has to be approximated. There is no inventory or materials handling, which rules out anything where physical stock is the constraint. And the published material does not describe fitting a statistical distribution to observed data, which matters more than it sounds: a model that assumes every step takes exactly eight minutes will systematically understate queueing, because the queue is created by the variation, not by the average.
The output is also a diagram and a table rather than an animated model. That is a presentation limit rather than an analytical one, and whether it matters depends entirely on who has to be convinced.
What the paid one buys
Simul8’s reputation is speed to a working model, and the reviewer consensus on that point is unusually consistent. It has a published self serve academy and short productised workshops, and the recurring theme in third party listings is that people become productive in it quickly.
Its browser version runs the same interface as the desktop application, which is a genuinely useful property: a model can be opened by somebody who has not installed anything. That is also its answer for anyone not on Windows, since there is no native Mac application, though the publisher’s system requirements also name virtualization software as a way to run the desktop product on a Mac.
It generates models from mined process data, which shortens the distance between the process as it really runs and the model of it. And it carries an animated model of the running process, which Bizagi does not.
The pricing problem
Simul8 does not publish a price. Third party listings report figures around five thousand per user, and separately six to eight thousand per user per year, which may describe different license terms. We have not been able to confirm either with the publisher, and we do not print unconfirmed figures as fact.
What we can say without an estimate is this: Simul8 offers no permanent free edition of any kind. Several of the products in our comparison set leave an evaluator something that does not expire. Simul8 offers a timed trial.
For a buyer with real interest and a budget that starts next quarter, that difference decides the matter on its own. The trial ends, the model closes, and the work stops.
How to choose between them
Start with Bizagi. It costs nothing, the documentation is public, and the four levels give you a clear ladder to climb. Build the model. If you reach level four and the answer you needed is in it, you have finished, and you have spent nothing.
You will know you have outgrown it when you hit one of three walls: you need to represent variation rather than averages, you need to model physical materials or batching, or you need somebody who does not build models to watch the process run before they will believe it.
At that point you are shopping for a discrete event tool, and Simul8 is a reasonable candidate on ease of use. Go into that conversation knowing you will not be told the price until you ask, and that you will not have a free copy to fall back on afterwards.