Home Assistant’s Voice Preview Edition can be a genuinely privacy-respecting voice device—but it is not automatically a fully local smart speaker. Its physical microphone-mute switch cuts power to the microphones, its firmware is open source, and wake-word detection can run on the device. But speech-to-text and text-to-speech still depend on the voice pipeline you choose: Home Assistant Cloud, local services on your own hardware, or a hybrid configuration.
That makes the Voice Preview Edition best understood as a dedicated Home Assistant voice satellite, not a standalone Alexa or Google Nest replacement.
As an Amazon Associate I earn from qualifying purchases.
What the Voice Preview Edition actually is
The Home Assistant Voice Preview Edition is a small voice-control endpoint designed and built by Nabu Casa, Home Assistant’s commercial partner. It provides the microphones, speaker, physical controls and audio hardware; a separate Home Assistant installation does the meaningful assistant work.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Clear out junk files and repair common Windows errors3Scan for outdated or missing drivers - takes under a minuteIn practice, the path looks like this:
Microphone → Voice Preview Edition → Home Assistant voice pipeline → local or cloud speech services → Home Assistant action.
#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.
It is based on ESPHome, ships with open-source firmware for its ESP32 and XMOS components, and is intended to be easier to deploy than a DIY microphone-and-speaker project. It is not a general-purpose smart speaker that works independently from Home Assistant. You must already run Home Assistant on another device, or add a suitable appliance such as Home Assistant Green.
Why its privacy claim is more credible than a software mute button
A physical microphone cutoff
The side-mounted mute switch physically cuts power to the microphones. That is materially stronger than an app setting that merely tells software not to use them: while the switch is engaged, the microphones cannot continue capturing audio because they are not powered.
The claim should still be kept narrow. The switch addresses microphone capture by this device while muted. It does not erase audio already processed, control other microphones in your home, or prevent a cloud service from receiving audio when the device is unmuted and your selected pipeline uses one.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
On-device wake words
The device supports local microWakeWord models, including “Okay Nabu,” “Hey Jarvis” and “Hey Mycroft.” The normal product configuration does not currently support custom wake words. Home Assistant’s wake-word documentation explains the available approach.
Local wake-word detection is useful for privacy, but it is not the same as fully local voice processing. Detecting “Okay Nabu” on the device does not prove that the sentence spoken afterward will be transcribed locally.
Open firmware, without overpromising
Home Assistant’s product materials describe the ESP32 and XMOS firmware as fully open source, with ESPHome preloaded. That supports inspection, modification and community development. It is not a blanket security certification, and open firmware alone does not determine whether your particular setup sends audio to the cloud.
Nabu Casa also positions the product within Home Assistant’s Open Home philosophy rather than an advertising-based smart-speaker model. That is useful context, but it should not be confused with a guarantee that no data can ever leave your network.
“Local” can mean three different things
The most important buying distinction is whether “local” describes the wake word, the speech processing, or the resulting home-control action.
| Configuration | Speech processing | Does audio leave home? | Typical difficulty |
|---|---|---|---|
| Home Assistant Cloud | Speech-to-text and text-to-speech use Nabu Casa cloud services | Yes | Low |
| Fully local pipeline | Speech-to-text and text-to-speech run on your Home Assistant hardware or local services | No, assuming no external integrations are involved | Medium to high |
| Hybrid or LLM-backed setup | Depends on the local and paid services selected | Depends | Variable |
Home Assistant Cloud
This is the simplest route. The Voice Preview Edition captures your command, audio is sent to Home Assistant Cloud for speech processing, and the resulting action is handled through your Home Assistant instance. Turning off a light may therefore involve cloud transcription while the actual smart-home command still executes through your local installation.
Cloud is often the better-performing and better-supported option for modest hardware or languages with limited local-model support. It is not fully local because voice audio leaves your home network. See Home Assistant’s cloud voice assistant documentation for the processing details.
Fully local speech processing
A local pipeline keeps speech-to-text and text-to-speech on your own hardware or local services and does not require a Home Assistant Cloud subscription. The trade-off is configuration, processing power and language coverage.
Home Assistant recommends at least an Intel N100-class processor for a reasonably responsive Whisper Base setup in supported scenarios. A lightweight Speech-to-Phrase pipeline is more efficient for focused home-control commands, but it is deliberately constrained and is not equivalent to general-purpose dictation. Whisper is more flexible, but larger models need more capable hardware.
“Local” is also language-dependent. A language may work well through Home Assistant Cloud while having weaker local models, slower recognition or no practical local text-to-speech option. Whisper may technically support a language without delivering the accuracy you expect.
Hybrid and AI-enhanced configurations
Assist can use a local or paid large language model as a conversation agent or fallback for requests that normal Home Assistant commands do not understand. This creates a separate privacy and cost decision. A local LLM can preserve the local processing model; a hosted LLM may send prompts or transcripts to another provider.
What is in the device?
The Voice Preview Edition has a dual-microphone array, an internal speaker, a 3.5 mm stereo headphone output and 2.4 GHz Wi-Fi. It also supports Bluetooth 5.0 and uses USB-C power at 5 V DC, 2 A.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →- Rotary dial: adjusts volume and provides other input.
- Center button: can activate the device without a wake word and cancel a response.
- LED ring: indicates listening, processing and responding states.
- Mute switch: physically cuts power to the microphones.
The unit measures 84 × 84 × 21 mm and weighs 96 g. It includes a Grove port, exposed PCB pads and a screw-open enclosure for expansion and experimentation. It is intended for indoor use from 0–30 °C in non-condensing humidity.
Important: the box does not include a USB-C charger or cable. You need a suitable power supply separately.
What it can control
The device can control Home Assistant entities exposed to Assist. That can include lights, switches, fans, covers, media devices and thermostats. Depending on your configuration, it can also run automations or scripts and manage features such as shopping lists and timers.
Rank #3
Its capability depends on more than the hardware. You need:
Free tools Windows power users keep installed
One-click scans. No signup required.
- a device Home Assistant can control;
- an entity exposed to Assist;
- a clear entity name and correct area assignment;
- a supported command and language;
- a functioning voice pipeline and conversation agent.
Expose only the entities you want voice control to access. Short, distinctive names and sensible areas make commands more reliable. Duplicate names, unusual punctuation and ambiguous room assignments can make a good pipeline appear unreliable.
What it does not replace
The Voice Preview Edition is primarily for smart-home control. Home Assistant itself does not present it as a complete replacement for every feature of Alexa, Google Assistant or Siri.
It is not the strongest choice if your priority is:
- Alexa’s broad skills ecosystem;
- Google Assistant’s general web knowledge;
- Siri’s deep Apple-device and Apple-service integration;
- music-focused speaker performance and media services;
- shopping, broad factual questions or open-ended conversation;
- a hands-free chatbot that understands almost any request.
It can replace an Echo for many commands such as “turn off the kitchen lights,” but that is a narrower claim than replacing an Echo as a general household assistant.
How setup works
Menu labels can change between Home Assistant releases, but the stable setup path is:
- Run Home Assistant on a supported computer, appliance or home server.
- Connect the Voice Preview Edition to USB-C power.
- Make sure a 2.4 GHz Wi-Fi network is available.
- Add the device through Home Assistant’s device setup flow.
- Connect it using the setup wizard.
- Select or create an Assist voice assistant.
- Choose Home Assistant Cloud, a local pipeline, or a hybrid configuration.
- Select an available wake word.
- Assign the device to a Home Assistant area.
- Expose the devices and entities you want Assist to control.
- Test both wake-word and physical-button activation.
- Engage the mute switch and verify that microphone input stops.
For a privacy-focused installation, do not stop after adding the hardware. Confirm which pipeline is active, which services it uses, what language is selected and whether any optional LLM or cloud integration is enabled.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.If setup or recognition fails
Check the basics first:
- Confirm the device has adequate USB-C power.
- Verify that 2.4 GHz Wi-Fi is available and the signal is adequate.
- Check communication across VLANs, guest networks and firewall rules.
- Update Home Assistant and confirm that the selected pipeline is installed and running.
- Check local speech-to-text and text-to-speech support for your language.
- Make sure the microphone switch is not engaged.
- Confirm the device is assigned to the intended area.
- Rename ambiguous entities and expose the correct devices to Assist.
If a local setup is slow or inaccurate, temporarily compare it with Home Assistant Cloud. If Cloud works while local processing does not, the likely cause is local hardware, model selection, language support or service configuration rather than the Voice Preview Edition itself.
Placement matters too. Keep the unit in the open rather than behind objects or in a corner. Distance, reflections, television and music volume, room layout and accent all affect recognition. Wake words can also trigger accidentally, which is why the available choices are deliberately distinctive rather than ordinary words.
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 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchRank #4
- Fitment For: Designed for Home Assistant Voice Preview Edition, ensuring a secure and stable wall mounting fit.
- Optimized Voice Interaction: Positions your device at a convenient height for better voice control experience.
- Stable Device Support: Provides firm support to keep the device securely positioned during daily use.
- Space Saving Wall Mount: Wall mount holder keeps your smart speaker off desks and counters for a cleaner setup.
- Multi Room Application: Ideal for kitchens, bedrooms, offices, and other smart home locations.
Specifications and current price
Home Assistant’s current product page lists a recommended MSRP of $69 in the United States and €59 in Europe. Taxes, regional pricing and retailer pricing vary. The original December 2024 launch announcement cited a historical $59 recommended MSRP; the current product-page figure is the more relevant buying reference.
The device launched in December 2024 and remains labelled a Preview Edition. That status matters: software, UI labels, language coverage and capabilities may continue to evolve. It is better viewed as a supported platform for an improving open voice assistant than as a finished cloud-assistant replacement.
Alternatives
Use Assist through your phone
The Home Assistant mobile app with Home Assistant Cloud is the lowest-friction option if you do not need a dedicated room device. It avoids buying hardware, but it is less convenient for hands-free, room-by-room control.
M5Stack ATOM Echo
The M5Stack ATOM Echo is a low-cost voice-assistant development option documented by Home Assistant. It suits makers and experiments, but it is more development-board-like and lacks the Voice Preview Edition’s enclosure, dial, larger speaker presentation and dedicated physical microphone cutoff.
Seeed Studio and DIY ESPHome satellites
Seeed Studio sells ReSpeaker and other voice-development hardware, while Home Assistant documents compatible ESPHome satellite approaches. These routes can offer lower cost or more customization, but they require more work around flashing, microphones, enclosures, power and troubleshooting.
Alexa, Google Nest and Apple HomePod
Conventional smart speakers remain stronger for minimal setup, music, general questions, broad third-party ecosystems and mature far-field behaviour. The Home Assistant device is the more compelling choice when control over the processing path, open firmware and a physical microphone cutoff matter more than general-assistant breadth.
Who should buy it?
Buy it if you already run Home Assistant, want a polished voice satellite, care about a genuine hardware microphone cutoff, mainly issue home-control commands and are comfortable configuring entities, areas and voice pipelines. It is also a good choice if you want the option to start with Cloud and move toward local processing later.
Skip it if you want a plug-and-play standalone speaker, primarily want music or general questions, expect Alexa-level ecosystem breadth, need custom wake words now, have weak hardware and do not want Cloud, or need a charger included in the box.
Recommended Free Tools
The hardware price is only part of the decision. You may also need an existing or new Home Assistant server, a USB-C power supply, more capable hardware for local Whisper processing, a Home Assistant Cloud subscription, additional satellites or optional LLM services.
Quick Recap
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.
