If you're less concerned about privacy, I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant while being much cheaper than the electricity that would be needed to keep a 3090 awake. The thing that kills this for me (and they even mentioned it) is wake word detection. I have both the HA voice preview and FPH Satellite1 devices, plus have experimented with a few other options like a Ra…
My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
101–110 of 153 posts
Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#102Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#103If you're less concerned about privacy, I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant while being much cheaper than the electricity that would be needed to keep a 3090 awake. The thing that kills this for me (and they even mentioned it) is wake word detection. I have both the HA voice preview and FPH Satellite1 devices, plus have experimented with a few other options like a Ra…
Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#104actually the hardest part of a locally hosted voice assistant isn't the llm. it's making the tts tolerable to actually talk to every day. the core issue is prosody: kokoro and piper are trained on read speech, but conversational responses have shorter breath groups and different stress patterns on function words. that's why numbers, addresses, and hedged phrases sound off even when everything else works. the fix is t…
80% of my home voice assistant requests really need no response other than an affirmative sound effect.
Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#105If you're less concerned about privacy, I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant while being much cheaper than the electricity that would be needed to keep a 3090 awake. The thing that kills this for me (and they even mentioned it) is wake word detection. I have both the HA voice preview and FPH Satellite1 devices, plus have experimented with a few other options like a Ra…
How about a button? I'd prefer to physically press a button on an intercom box than having something churning away constantly processing sound.
Or do you mean a button that activates chunked recording, passes it to a speech-to-text model, forwards to an LLM to infer intent, which triggers HA to issue a command, over a wireless network, to the computer with the light attached, to tell the light to turn on.
Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#106Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#107Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#108Is there a locally hosted voice assistant for Android phones? One available through F-Droid, if possible.
Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#109Re: My Journey to a reliable and enjoyable locally hosted voice assistant (2025)
#110Earlier quoted context omitted.
How about a button? I'd prefer to physically press a button on an intercom box than having something churning away constantly processing sound.
Rules out a bunch of cases where your hands are busy handling ingredients in the kitchen, etc