If you're comfortable where you are, then no. Personally, I use a mixture of both. For browsing through commit history or viewing diffs before committing / pushing, I find the GUI to be much more convenient. For most other things, I use the command line.

- easier to pick partial commits - easy visualisation of where branches split and merge - same easy visualisation of stashes

I use both the CLI and GUI. They both have their merits and demerits. While a CLI may have faster commands, a GUI helps a lot when trying to make sense of relationships between commits branches and visualisation in general.

The main benefit for me is the ability to quickly and conveniently see the commit history and diff as well as branches' current position. If you're already comfortable with doing the same in CLI, then you're probably not going to gain a ton.

I like Sourcetree for this particularly because its buttons map fairly well to the actual CLI commands that it will be executing, and makes it painless to open a CLI terminal for instances where I want more fine-tuned control.

As someone who made the opposite transition: no. My colleagues that are still GUI bound are generally slower at accomplishing most day to day tasks. Committing, rebasing, merging, pushing tags, etc.

For me the value is in diffing, merging and browsing the commit graph. Piecemeal staging can be improved looking at the whole file instead of the window `add -p` shows.

But... I'm still primarily a CLI user. I've just configured difftool and mergetool to call up Beyond Compare. git gui and gitk make designers weep, but I've never felt hindered or slowed down in what I want to accomplish with them.

I'll try a new GUI tool every now and then, and I haven't found any that make me feel more productive than the CLI augmented by the tools above.

I've found graph views and cherry picking to be better experiences in various guis, but generally for graphs I use tig or gitk and for cherry pick I suffer and copy paste commit hashes. The weight of the various guis and the obscurity of internals is usually enough to stop me from using it.

The only use I have found for a GUI is partial commits, and for that I use gitk which is already distributed with git.