I am in the first half of my PhD (Pure Mathematics) in the U.S. In this period of time I've read all of my advisors papers (apart from the very early ones), discussed all of them with him, and improved on many of them. My advisor is great: he has been extremely helpful, he'd always read all my stuff, and I definitely learnt a lot from him. I think he also got something from me, since I was able to answer a conjecture and a couple of questions that he had formulated/thought about for quite some time.
However, I feel like I know all the stuff he does at this point: he has very little left to teach me, as I noticed in the last six months.
My current situation is not too bad, I am capable of finding my own projects and most of my publications are based on my own ideas (I have 10 preprints/publications), so I could just go on like this for 2-3 more years. However, I think it would be better for me to go somewhere else, to learn from some other mentor, maybe in some different but related field.
Indirect sentences make me believe that my advisor does not want me to leave, although I never asked him directly. Before I do so, I would like to hear your opinion (I understand that the information I gave is not exhaustive, but nevertheless...).