Does this really work for anyone? I can sit there, right above, “hey nabo” over and over and maybe once in 20 times, it responds? The timer controls are terrible and that should be the easiest thing - if I setup two timers, and then ask for the status, most of the time it will say “I don’t know the device timer status”, and then occasionally it will say, “there are two timers. The five minute timer has five minutes.” So, it always says the five minute timer has five minutes, and it never mentions the other one.
The light is half on right now - “hey nabu, why is your light on?” “I’m not aware of any device called ‘why is your’”.
It kind of seems like this isn’t really even alpha yet - should I look at this again in 6 months?
Well, for it to answer when I talk to it, heh. Right now, that works VERY rarely - I have to hit the button for anything. If I could get that to work, then all I want is timers, alarms, running automations and turning on and off devices.
First voice. Wake word. That’s one problem. The wake work on device is good but can get better with tuning take that as a separate issue from what it (the device) does/can do. You can hit the button and it hears you.
Which wake word are you using - Nabu?
As for what it does. There are two kinds of voice resp systems in HA. One is speech to phrase. One is llm.
BOTH should be considered very basic to start but ultimately extensible. Out of the box expect it to be able to do an on device timer and turn basic things on off, set a thermostat and some basic volume. Beyond that you WILL need tools.
Are you using an LLM or basic assist speech to phrase.