We don’t need more models, improve on the Open Source ones we already have.
We need better hardware with long ranging microphones that can isolate the voice.
I’m still waiting for something I can put in my rooms, I can even make use of my Sonos speakers for audio output, so it’s really just the microphone.
Got the same setup, but I connect to my server with vscode through SSH and edit files there.
Much more user friendly as a primary Windows guy.