I’m a self taught Python dev and struggle to know when to use classes

There are 2 main cases. 1. there is a very clear "thing" in your code and the easiest model is how those "things" interact with each other. Like if you were writing a game like a Pacman clone, you can divide that easily into a class for Pacman and a class for ghosts. You have clear "things" that have data associated with them, your program is gonna be about their interactions, and it's a tidier design than trying to track all this data through lists or individual variables. Languages like C# make their whole language design like this but Python doesn't, so it's not always the most Pythonic way to do things. But if you reach this pattern, it's painfully obvious that this is what you should do. A good rule of thumb is that a function name should be a verb, and a class name should be a noun. Using inheritance or composition you can also model relationships between classes quite well. 2. You have a bunch of functionality that is related (even if not by an explicit thing) and you want to keep it together. It's a way of keeping related functions together, like how you would use namespaces in other languages. As an example, you might have a bunch of logic trying to handle a bunch of different errors, so you might choose to keep that together in an error handler class. It's also a really nice way of working with other people, since you can each take a job, write and test your code in a class each, then putting the thing together at the end is just deciding how everything fits together. It's like when I take your code, I just need to know what's happening without knowing the implementation details. I think a good takeaway is to go learn a language like C#, see how they do things over there, then come back to Python. C# forces you to write code in OO style and python doesn't. This way you can reach a better understanding of when it's the best thing to do.