The Execute Python Script Node
The preferred method for calling Python inside of an Editor Blueprint Asset is the "Execute Python Script" BP Node
Last updated
The preferred method for calling Python inside of an Editor Blueprint Asset is the "Execute Python Script" BP Node
Last updated
For general use, especially if releasing something to other users, it is recommended to use the Epic-provided Execute Python Script node. This is the most robust option, allowing us to run Python scripts and capture their output in the Blueprint Graph:
The advantage of this node is that we can declare its inputs and outputs in the Details panel, allowing us to pass wildcard data between Blueprints and Python:
The BP Graph does not track Python errors on its own, if any errors occur in our Python script the BP Graph will just continue on to the next node. The Success? default output allows us to abort an is a default return on the node, if any Python errors occur it will return False, allowing the opportunity to abort if needed:
Note: This feature was fixed in 5.5, it does not appear to work in 5.2-5.4 (tested December 2024)
A workaround for affected Unreal versions is to provide our own success variable:
Inputs and Outputs are added as wildcards by default on the node (colored grey). The BP Node relies on the Graph Pin connections to determine the variable type of each Input/Output.
For Inputs added in the Details Panel we will need to connect a Blueprint Graph Pin to provide their starting value + data type. Inputs are automatically initialized in the Python and are ready to use right away:
For Outputs added in the Details Panel we also need to connect a Blueprint Graph Pin, this tells the Execute Python Script what data type to use for each output:
The main advantage of this method is its reliability, as long as the Python code inside is valid there shouldn't be a problem. This is the preferred method to run Python code from a Blueprint.
A cautionary disadvantage of this method is code management, especially for larger productions/pipelines. It can be tedious to keep track of and update every instance of these nodes across multiple blueprints.
One recommendation to help reduce the management burden is to keep the bulk of your code in external Python modules and just have the Python Script node import & call functions:
A minor inconvenience is that the Inputs and Outputs are all wildcards relying on Pin connections to inform them of their data types. In some cases you may find yourself adding arbitrary nodes or extra variables just to declare the type:
