Protected Model Version Compatibility across MATLAB Releases and Updates

When I attempt to simulate a protected model created in R2017b General Release on a machine with R2017b Update 9, I get the following error:

ERROR: The protected model was created with an incompatible version of slprj. Recreate this
protected model with the version of Simulink that you are currently using and try again.

Does this mean that protected models are not compatible between updates within a release? How will this be handled by Simulink for future releases?

Answer 1

[MathWorks Support Team]

For detailed information and limitations for protected model simulation and code generation capabilities across MATLAB releases, refer to the following documentation:

Use Models Protected in Previous Releases

Protected model version compatibility can be summarized as follows:

• Protected models created in R2018b or later releases are forward-compatible. This means that the protected model can generally be simulated with any update version within the same release, or in a later Simulink release. Support for using protected models from previous releases in Just-in-Time Accelerator Mode was added in R2020a, and for Code Generation/SIL/PIL modes in R2020b.
• Protected models created in R2018a or earlier releases are not forward-compatible. You must use the exact same Simulink release and update version that was used to create the protected model. The protected model report will show the release versions used during its creation.

Additional Notes:

1. Version Compatibility: Note that the protected model should be created within 10 releases of your current version to ensure compatibility. This translates to a timeframe of 5 years. While models older than this might still function, we do not test them for compatibility, and thus cannot guarantee their compatibility or performance. 

2. Simulation: To simulate a protected model, make sure you set up MATLAB with the same compiler (via "mex setup") that was used to create the protected model. This is because protected models are stored as compiled MEX binaries and rely on the runtime libraries associated with that compiler to function correctly.

3. Code Generation: Generating code may require the same MATLAB release and update level (see this MATLAB Answer). You might also need to use the same toolchain as specified by the protected model creator.

4. Backward Compatibility: Protected Models are not backward compatible, meaning you cannot use a protected model created in a newer release than your current release. For example, a model from R2025a won’t work in R2024b. 

5. Known Bug in R2019b-R2021a: There is a known bug affecting compatibility between update versions within a release for R2019b through R2021a. For more details, refer to this MATLAB Answer.