I'm sharing the model I used to test Gemini 3.1 Pro writing FlexScript, as part of an ongoing series benchmarking LLMs for FlexSim.
The full article with context, prompts, and detailed analysis is here: I Tested Gemini 3.1 Pro on FlexScript. Here's What Actually Happened.
What’s in this model?
- A simple model finishing with two Queues for visual purposes.
- Triggers configured for the 6 tasks at Levels 1 and 2 described in the article:
- L1.1 – Get the current object’s name (OnEntry of Queue1)
- L1.2 – Read the priority label from the incoming item (OnEntry of Queue1)
- L1.3 – Print the current number of items in the queue (OnEntry of Queue1)
- L2.1 – Set entryTime label with the current simulation time (OnEntry of Queue1)
- L2.2 – Route items by priority on the Processor (Send To Port trigger)
- L2.3 – Delay item release from the queue based on the global variable releaseDelay
Each trigger contains the script generated by Gemini 3.1 Pro exactly as shown in the article, so you can:
- Inspect the code directly in FlexSim
- Run the model and see the behavior in the Output console
- Compare it with your own corrections or alternative implementations
How to use this model
- Open the attached .fsm file in FlexSim (version 2026 or later recommended).
- Run the simulation and watch the Output console to see each task’s results.
- Open the relevant triggers (OnEntry, Send To Port, etc.) to review the FlexScript generated by Gemini.
