Hey guys,
I received my Nabu Casa Home Assistant Voice Hardware today and I was really excited to set it up.
However I was disappointed very quickly when I found out, that I cannot use local speech-to-text and text-to-speech options.
Is there any way to enable this without the add-on store?
I’m running Home Assistant Core on my Raspberry Pi 5 with Arch Linux and I don’t need / want a supervised installation for anything else …
Will Whisper / Piper be included in HA Core anytime soon?
What else can I do? Can I install these manually somehow?
Hi, welcome to the forum!
There are quite some topics here about using Core with TTS/STT.
I’m also running Core with whisper/piper in docker on another host.
About whether it will be included in Core: most likely not.
Thanks. Do you know any helpful topics for my specific issue?
I have a native installation of HA Core from Arch Linux package repository …
No, not right away.
I had to do some search to get it working.
Did you search here or on your favorite search engine?
You simply run them as Wyoming services, with Docker or (if you’re into that kind of thing) directly on the OS:
Then you go to Settings > Integrations, add Wyoming Protocol. Then in Wyoming Protocol you click Add Service and point it to localhost port 10300 and then again with port 10400. Then under Voice Assistants you configure your pipeline to use them as STT and TTS.
1 Like