Alexa+ is Amazon’s new premium voice assistant that costs $20 per month (but comes free with Amazon Prime). Amazon’s top-rated feature is a conversational assistant that’s much chattier than the previous version.
With Home Assistant, you can achieve something similar for free.
Alexa+ style conversations with Home Assistant
Use LLM to chat and interact with your smart home
Amazon’s flagship evolution, Alexa+, is a more “conversational” assistant. This has been achieved with the use of a large language model (LLM), which essentially merges the chatbot style of ChatGPT or Gemini with a smart home voice assistant that controls your smart home.
Alexa+ is more than just this (and we’ll get to that later). But if what you want is to have chatbot-style interactions using a Home Assistant voice assistant, you’re in luck. You can even choose to use cloud-dependent models (what is Alexa+) or run your own models locally and keep everything offline.
I have two voice assistants on my Home Assistant server right now. One uses Google services (including Gemini) for conversation, text-to-speech (TTS), and speech-to-text (STT); the other uses completely offline models for everyone. Gemini is faster and easier to set up, while the offline models required a little more tinkering but still work consistently.
Add Google Gemini to Home Assistant
To add Gemini to Home Assistant, go to Google AI Studylog in, click Get API keyand copy it to your clipboard. Now go back to Home Assistant and click Settings > Devices followed by Add integration in it Integrations eyelash.
Search for “Google Gemini” and then paste the API key you retrieved earlier into the box. Hit Deliver and then Jump and finish to complete the setup. Alternatively, you can do a similar process with integrations for OpenAI, Anthropic, and other cloud-based LLMs.
You will only receive a limited number of API calls with a cloud provider per day. Although this is technically free, you will have to pay once you run into its limitations.
Create a local AI assistant with Ollama and the Wyoming Protocol
If you want to do your own LLM, you can do that. Be is a solid starting point, so Download it and install it on any server you are going to use.. From here, go to Settings > Integrations and install the “Ollama” integration. Point the integration towards the local IP address where Ollama is running (for example, http://192.168.0.149:11434).
You will have to figure out which model you want to use with Ollama and download it. I personally decided to qwen2as it’s small and fast enough that my modest M1 Mac mini with 16GB of RAM can handle it without drowning. I tested this on my server by running ollama run qwen2 in Terminal (which forced Ollama to download the model), then launch the app and have a simulated conversation in a new chat.
Now in the Ollama integration click on Add conversation agent and create an agent using the chosen model. You can edit the message, enable Home Assistant control (which you probably want to do), and more.
Lastly, you will also need some means to handle TTS and STT locally. The Open Home Foundation’s Wyoming Protocol is perfect for this. Go to Settings > Devices and services and install the “Wyoming Protocol.”
Next, go to Settings > Applications and download “Piper” and “Whisper” and then restart your Home Assistant server. Make sure both Piper and Whisper are running and return to the Devices and services screen. Home Assistant should automatically detect these services and give you the option to “Add” both. Almost finished!
Building your assistant and controlling your home
everything is connected
From here, it’s a simple case of heading to Settings > Voice assistant and creating a new one (or editing the one you already have). You can choose a chat agent from the drop-down menu, as well as STT and TTS models.
From here, you can use your voice assistant to chat as you would with Alexa+. You can do this by activating your Home Assistant smart speaker and asking a question, or by clicking the “Assist” button on the web UI panel or in the smartphone and smartwatch apps. You can even build or convert your own smart speaker using an ESP32.
In addition to chaining commands like “turn off the kitchen lights and close the front door,” you can ask your assistant all kinds of questions related to the chatbot. Just remember that this is an LLM and is subject to the same weaknesses as any other. He willpower Confidently answer incorrectly when asked how many R’s are in the strawberry!
Both methods allow you to customize the message used. This is what is sent to the model of your choice before each consultation. You can use this to add unnecessary personalityor create a convenient assistant that always responds with short sentences.
There is also a “Think before you answer” option for local LLMs, if you prefer better quality answers at the cost of speed.
Alexa+ wins in personalization and service connectivity
But it’s a privacy nightmare.
Alexa+ is much more than just the LLM integration, but this is arguably one of the biggest draws for many. Conversational style means you can continue chatting with Alexa+ without saying the wake word again. Home Assistant already has this to the extent that it will anticipate a response if it asks you a question, but it’s not at the level of Alexa+.
The same goes for personalization and learning. Alexa+ can learn about your allergies and food preferences, while remembering which route you took on your last road trip. Home Assistant can reference your calendar, but only if you use or import your own schedule into the platform’s underused calendar element.
Alexa+ talks to external services. You can buy things on Amazon, reserve a table at a restaurant, or order an Uber just by asking Alexa+. Home Assistant can’t do any of these things, but that’s not necessarily a bad thing. For many, the privacy implications of your smart home assistant knowing where you live, what you eat and where you go will be too much to bear.
-
- Dimensions
-
84x84x21mm
- Weight
-
96 grams
Home Assistant Voice Preview Edition is a privacy-first smart speaker created as an alternative to Amazon Alexa and Google Nest Mini. Adds voice assistant capabilities, including local-only processing, to a smart home powered by Home Assistant.
-
- Brand
-
Seed study
- UPC
-
ESP32-S3R8
The reSpeaker Lite voice assistant kit includes a dual-microphone array, a pre-soldered XIAO ESP32-S3 controller, and an XMOS XU316 audio processor with built-in natural language understanding, interference cancellation, acoustic echo cancellation, noise suppression, and automatic gain control. With a 5W speaker connected, you can create your own local voice assistant that you can connect to Home Assistant via ESPHome.
Alexa+ will appeal to those who are truly entrenched in the Amazon ecosystem. For me, nothing beats Home Assistant, especially a Cloud-dependent service like Alexa that could put your entire smart home at risk.
