There is a persistent assumption that Home Assistant is the powerful-but-dumb option — great at execution, no intelligence. That has not been true for a while, and repeating it does a disservice to anyone trying to make a decision.
Home Assistant can run a large language model as its conversation agent, entirely on your own hardware, today. It is worth understanding properly, including if you are evaluating us.
What works now
Local LLMs as the conversation agent. Home Assistant ships a native Ollama integration, so a model running on your own machine can act as the conversation agent — no custom glue code. You say "make the living room cosy" and the model maps that into concrete device calls: dim the lights, warm the colour temperature, lower the blinds. It is not keyword matching. The model can weigh time of day, who is home, and current sensor states against a natural-language goal.
A genuinely local voice pipeline. The Assist pipeline chains speech-to-text, the conversation agent, and text-to-speech over the Wyoming protocol, an open standard from the Rhasspy project. With Whisper for transcription and Piper for speech, the whole loop runs in your house.
Speed that is actually usable. Home Assistant reworked its text-to-speech to stream: Piper and Home Assistant Cloud TTS can begin speaking as soon as the model produces its first few words, which they measured as roughly a tenfold improvement in spoken response time. On a GPU with 8 GB of VRAM, a 7–8B model typically responds in under two seconds.
Reasonable model choices. Community consensus in 2026 has settled around the Qwen3 family — roughly Qwen3 4B Instruct on 8 GB cards, Qwen3-VL 8B in the middle, Qwen3 30B on 24 GB cards.
That is a real local AI stack, free, and open source. Credit where it is due.
What it costs you to get there
The capability is real. The path to it is not free of effort.
Hardware. A responsive local LLM wants a GPU. A Raspberry Pi running Home Assistant will not do this well. That is a second machine, its power draw, and its noise.
Assembly. Ollama, a model choice, Whisper, Piper, the Assist pipeline, and the exposure settings that decide which entities the model may touch. Every piece is documented; there are still a lot of pieces.
Model behaviour. The genuinely hard part is picking a model that calls Home Assistant's tools correctly instead of guessing at them. Smaller models hallucinate device names and invent capabilities. This is tuning work.
Maintenance. Models improve, integrations change, and what you assembled in spring wants revisiting by autumn.
None of this is a criticism. It is the normal cost of an open, composable system, and for a lot of people it is a pleasure rather than a chore.
Curious whether Nexop fits your home?
Book a live demo — one of the founders runs it, not a salesperson — or join the waitlist and hear from us the day it ships.
The gap that remains
Here is the part worth being precise about, because it is the only place we genuinely differ.
Everything above is reactive. However sophisticated the model, the interaction starts with you. You ask; it acts. A local LLM in Home Assistant is a much better interface to your automations. It does not reduce how many automations you have to think of.
Nothing in that stack watches a week of your household and says you dim the lights and close the blinds around 20:30 most weeknights — want that automatic? You still supply the intent. The model is a superb translator of intent, not a source of it.
That is not a flaw in Home Assistant. It is a different product goal. But if your problem is "I never got around to building the automations", a better command interface does not solve it — that is the proactive automation problem, and it is a different piece of engineering.
Which should you choose
Stay with Home Assistant if you enjoy assembling systems, you already have hardware that can host a model, or you want maximum control with no vendor in the loop. It is free, it is excellent, and the ceiling is far above ours. If you want the AI parts without the assembly, Home Assistant Cloud at $6.50/month is a fair deal that also funds the project.
Consider Nexop if you want the local-first properties without building the stack, and specifically if you want the house to propose automations rather than wait to be asked. We support Zigbee, KNX, Matter, and WiFi natively — no dongle needed for Zigbee — but note that Home Assistant's integration library is vastly larger than ours will be, and we do not support Z-Wave or Thread at launch.
The fuller side-by-side is on our Home Assistant comparison, which also says plainly where Home Assistant is the better choice.
Nexop is pre-launch. If you want a local AI voice assistant this month, the Home Assistant stack above genuinely works and we would rather point you at it than pretend otherwise. If the proactive part is what you have been waiting for, book a demo or join the launch list.