I have tried several offerings, including WisprFlow, which many friends recommended and use, as well as Apple's own tools. But I was reluctant because (a) I don't want to pay, and (b) I don't want my voice going to some company capturing my ideas/voice forever.
So this just uses your Mac's local hardware, running the model locally, and it's all yours, all private, all open source. Thoughts/contributions welcome.