That's actually a bit of a design issue at the moment. I don't have a good way to use a keyboard when in VR and have to use the push-to-talk button when chatting to the companion. I also have to keep the voice messages short, or long audio clips run into processing issues for some reason. Lots of issues!
However, it's a design oversight on my part that this doesn't help you with inputting the main-menu details. I need to add the ability to use a local LLM and speech-to-text more cleanly, for sure.
Thanks for the kind words and the feedback!