The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →A local LLM can make Home Assistant voice control feel more flexible, but that does not make it the best choice for every request. In my setup, I tried an LLM as a voice conversation agent, then turned it off for most of what I say. The useful distinction is between routine commands, which Home Assistant can handle with built-in intents, and open-ended requests where an LLM’s conversational flexibility may help.
Why I stopped routing most commands through the LLM
My experience was that I did not need an LLM for most of the things I said to Home Assistant. Routine requests—such as controlling a light—are a better fit for predictable intent handling than for a general-purpose model. I kept the LLM as an option for language that benefits from a less rigid, more conversational interpretation.
As an Amazon Associate I earn from qualifying purchases.
That is a personal choice, not evidence that local LLM voice control is inherently unreliable. Home Assistant’s documentation does, however, describe its Ollama control feature as experimental and identifies constraints worth considering before making it the default for every utterance.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWhere the LLM fits in Home Assistant voice control
Assist is a pipeline, not a single model. Speech recognition turns audio into text; a conversation agent interprets that text; Home Assistant executes an intent or action; and text-to-speech can speak a response. The microphone or satellite, speech-to-text service, conversation agent, and speech output are distinct parts of the system. An LLM changes the conversation-agent step—it does not automatically replace the other components. Home Assistant’s voice overview explains these parts.
#1 Best Overall
- Designed for Home Assistant Voice & Music Workflows: Preloaded with Home Assistant Voice Assistant and Music Assistant. Functions as both a voice input terminal and an audio playback endpoint.
- Dual Microphones for Voice Capture: Built with dual digital microphones for wake word or button-activated voice capture. Audio is streamed to the Home Assistant voice pipeline.
- Integrated 3W Speaker for Direct Playback: The built-in 3W/4Ω speaker supports TTS playback, Music Assistant streaming, and system audio without external speakers.
- Linux-Based Local Operation: Runs a lightweight Linux system on a quad-core ARM A53 CPU with 256MB RAM and 512MB flash for local audio processing.
- Development & Debugging Capabilities: Supports firmware flashing, and also provides access to live logs, on-device editing—suitable for routine development or issue diagnosis.
Home Assistant’s built-in conversation agent matches text to intents. Integrations can provide other agents. With the Ollama integration, Home Assistant connects to a separately running local Ollama server; when control is enabled, the model can use Home Assistant tools to get information about and control only the entities exposed to it. The Assist API exposes intents and the entity capabilities available to the built-in conversation agent, not administrative tasks. See the Ollama integration documentation and the intent conversation API documentation.
Built-in intents and an LLM solve different problems
| Consideration | Built-in intent handling | LLM conversation agent |
|---|---|---|
| Routine home commands | Matches text to supported intents, making it a natural fit for predictable requests. | Can interpret requests through tool calls, but adds a model and tool-support dependency. |
| Open-ended wording | Works within the intents and phrases Home Assistant recognizes; custom sentences and intents can extend explicit coverage. | May be useful when a request is less structured or conversational, subject to the model’s capabilities and reliability. |
| Entity access | Uses Home Assistant’s built-in intent and entity capabilities. | Can control only entities exposed to the agent; Home Assistant recommends exposing fewer than 25 while experimenting. |
| Sentence triggers | Can use local sentence handling and custom sentences. | The Ollama integration does not integrate with sentence triggers. External agents use sentence triggers only when “Prefer handling commands locally” is enabled. |
The table describes documented system behavior, not a universal verdict on which agent will understand a particular household best. Home Assistant’s Conversation integration documentation covers the local-command preference and external agents.
Rank #2
- 🎙️ Hands-Free Voice Typing for Windows & Mac – Powered by iOS & Android dictation technology, AI VoiceWriter allows fast, accurate speech-to-text directly on your desktop. Simply speak, and your words appear in real time. Compatible with Windows 10 & above, macOS 13 & above.
- ✍️ AI Writing Assistant for Effortless Editing – Boost productivity with AI proofreading, rephrasing, and formatting. Perfect for emails, reports, creative writing, and professional content.
- 💻 Works Seamlessly in Any Desktop App – Type with your voice in Microsoft Word, Google Docs, PowerPoint, Teams, emails, and more. Just place your cursor in any text field and start speaking!
- 📱 Mobile App for Enhanced Voice Input – The AI VoiceWriter mobile app enhances voice recognition by using your phone’s microphone as an input device for clearer, more accurate dictation—while typing on your desktop. Supports iOS 15 & above, Android 9.0 & above.
- 🌎 Multilingual Voice Typing & AI Assistance – Supports 33 languages for dictation, plus AI-powered features in Chinese, English, Japanese, Korean, French, German, Spanish, Italian and, Swedish.
What Home Assistant warns about with Ollama control
Home Assistant labels control through its Ollama integration experimental. Its documentation says only models that support tools can control Home Assistant, recommends exposing fewer than 25 entities while experimenting, and warns that smaller models are more likely to make mistakes. It also cautions that smaller models may not reliably maintain a conversation when control is enabled. These are official cautions, not proof that every local setup will fail.
Recommended Free Tools
- Keep the exposed surface narrow. Expose only the entities the agent needs, especially while testing. The fewer-than-25 recommendation is Home Assistant’s guidance for experimentation, not a guarantee of safe or accurate behavior.
- Check tool support. A model that cannot use tools cannot control Home Assistant through this integration.
- Set expectations for dialogue. A model may handle a one-shot command differently from a multi-turn exchange; the documentation specifically warns about conversation reliability with smaller models.
Home Assistant also documents using two Ollama configurations with the same model but different prompts: one for conversation without control and another with control. That approach can separate open-ended conversation from requests that need access to home entities. The integration page describes the option.
Rank #3
- Your favorite music and content – Play music, audiobooks, and podcasts from Amazon Music, Apple Music, Spotify and others or via Bluetooth throughout your home.
- Alexa is happy to help – Ask Alexa for weather updates and to set hands-free timers, get answers to your questions and even hear jokes. Need a few extra minutes in the morning? Just tap your Echo Dot to snooze your alarm.
- Keep your home comfortable – Control compatible smart home devices with your voice and routines triggered by built-in motion or indoor temperature sensors. Create routines to automatically turn on lights when you walk into a room, or start a fan if the inside temperature goes above your comfort zone.
- Do more with device pairing – Fill your home with music using compatible Echo devices in different rooms, or create a home theatre system with Fire TV.
- Say goodbye to drop-offs and buffering - With eero Built-in, Echo Dot doubles as a mesh wifi extender, adding up to 1,000 sq. ft. of wifi coverage to your existing eero network.
Local speech recognition is a separate decision
Choosing an LLM does not determine how quickly or accurately Home Assistant turns speech into text. Home Assistant’s local voice guide describes two different speech-to-text approaches:
- Speech-to-Phrase is a closed-ended model that transcribes what it knows and supports a subset of Assist commands. Home Assistant reports under one second on Home Assistant Green or Raspberry Pi 4. Its constrained command coverage suits home control more than open-ended transcription.
- Whisper is open-ended. Home Assistant reports around eight seconds on Raspberry Pi 4 and under one second on an Intel NUC. The guide recommends it for more powerful hardware when a household wants to extend voice beyond simple home control, including pairing it with an LLM.
Those timings are Home Assistant’s published examples, not guarantees for every device, language, or configuration. The same guide says performance and speech quality vary by device and language. It describes Piper as local neural text-to-speech optimized for Raspberry Pi 4 and reports that medium-quality models can generate 1.6 seconds of voice in one second on a Raspberry Pi. Speech recognition, conversation processing, and speech output each contribute to the experience; changing only the LLM does not remove delays elsewhere in the pipeline. Home Assistant’s local voice guide covers these options.
Latency and accuracy figures need context
A 2025 study by Rune Birkmose, Nathan Mørkeberg Reece, Esben Hofstedt Norvin, Johannes Bjerva, and Mike Zhang examined fine-tuned on-device LLMs for Home Assistant. It reported approximately 80–86% accuracy on noisy human prompts and out-of-domain intents, with average inference time of 5–6 seconds per query. The authors described that latency as acceptable for one-shot commands but suboptimal for multi-turn dialogue. Those measurements apply to the study’s models, tasks, and evaluation; they are not a benchmark for every Ollama model or for my setup. The paper provides the study details.
How to decide what should reach the LLM
- Start with the requests you actually make. If most are direct device commands, try built-in intent handling for those first rather than assuming every utterance needs a general-purpose model.
- Identify the gap. Reserve LLM handling for requests whose wording or conversational context is not well served by the intents you use. Custom sentences and intents can cover phrases where you want explicit local behavior.
- Limit access before enabling control. In the Ollama integration, expose only the entities needed for the experiment, and follow Home Assistant’s recommendation to expose fewer than 25 while experimenting. Confirm that the model supports tools.
- Evaluate the whole pipeline. Consider speech recognition coverage and latency on your host, the agent’s response time, and speech output. A delay may come from transcription or speech generation rather than the LLM alone.
- Keep control and conversation separate if useful. Home Assistant documents separate Ollama configurations with different prompts for conversation and control. Decide which types of requests should have access to home entities.
- Retain a reliable path for routine commands. Prefer local handling where a sentence trigger or explicit custom intent should take precedence; external agents use sentence triggers only when “Prefer handling commands locally” is enabled.
What an all-local voice setup means
Home Assistant describes a fully local pipeline in which audio is captured, transcribed locally, interpreted by Home Assistant, and spoken back by local text-to-speech. That depends on configuring every component locally; using a local LLM by itself does not establish that the entire voice path stays local. Home Assistant puts it plainly: “Your spoken commands never leave your home: a microphone hears you, a local speech-to-text engine turns your voice into text, Home Assistant figures out what you want, and a local text-to-speech engine speaks the answer back.” The local voice setup guide explains the full arrangement.
Quick Recap
Best Value
- Your favorite music and content – Play music, audiobooks, and podcasts from Amazon Music, Apple Music, Spotify and others or via Bluetooth throughout your home.
- Alexa is happy to help – Ask Alexa for weather updates and to set hands-free timers, get answers to your questions and even hear jokes. Need a few extra minutes in the morning? Just tap your Echo Dot to snooze your alarm.
- Keep your home comfortable – Control compatible smart home devices with your voice and routines triggered by built-in motion or indoor temperature sensors. Create routines to automatically turn on lights when you walk into a room, or start a fan if the inside temperature goes above your comfort zone.
- Do more with device pairing – Fill your home with music using compatible Echo devices in different rooms, or create a home theatre system with Fire TV.
- Say goodbye to drop-offs and buffering - With eero Built-in, Echo Dot doubles as a mesh wifi extender, adding up to 1,000 sq. ft. of wifi coverage to your existing eero network.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
