What are the differences between git remote prune, git prune, git fetch –prune, etc

I don’t blame you for getting frustrated about this. The best way to look at is this. There are potentially three versions of every remote branch:

1. The actual branch on the remote repository
   (e.g., remote repo at https://example.com/repo.git, refs/heads/master)
2. Your snapshot of that branch locally (stored under refs/remotes/...)
   (e.g., local repo, refs/remotes/origin/master)
3. And a local branch that might be tracking the remote branch
   (e.g., local repo, refs/heads/master)

Let’s start with git prune. This removes objects that are no longer being referenced, it does not remove references. In your case, you have a local branch. That means there’s a ref named random_branch_I_want_deleted that refers to some objects that represent the history of that branch. So, by definition, git prune will not remove random_branch_I_want_deleted. Really, git prune is a way to delete data that has accumulated in Git but is not being referenced by anything. In general, it doesn’t affect your view of any branches.

git remote prune origin and git fetch --prune both operate on references under refs/remotes/... (I’ll refer to these as remote references). It doesn’t affect local branches. The git remote version is useful if you only want to remove remote references under a particular remote. Otherwise, the two do exactly the same thing. So, in short, git remote prune and git fetch --prune operate on number 2 above. For example, if you deleted a branch using the git web GUI and don’t want it to show up in your local branch list anymore (git branch -r), then this is the command you should use.

To remove a local branch, you should use git branch -d (or -D if it’s not merged anywhere). FWIW, there is no git command to automatically remove the local tracking branches if a remote branch disappears.