The total number of commits and the maximum history depth. Cloning a repository with a large number of commits and a deep history and running certain local Git commands on it (e.g. git blame) can be slow.

The total size of blobs in MiB. Avoid a rapid growth of the total size of blobs by using LFS for larger files.

The total number of branches and tags. Cloning a repository with a large number of branches and tags can take a lot of time and local Git operations (e.g. tab-complete a branch name) can be slow.

The total number of submodules. Git needs to query their status on all kinds of local operations and this query operation is in particular slow on Windows.