When you first open HomeGenie, the Setup Wizard guides you through choosing your initial setup. You can explore demo devices, connect hardware controllers, or restore a previously saved configuration. You do not need to decide everything right now: the Settings section lets you enable additional features at any time.
If you have not downloaded and launched HomeGenie yet, see the Getting started guide to prepare your system.
Protect access to your server
Complete initial setup locally or on a restricted, trusted network. Completing the Setup Wizard does not configure a password on its own. Before connecting other client devices, open the HTTP settings in the Maintenance menu, turn on the Password protected access option, enter a strong password, and click the Confirm button. By default, the username is "admin".
Check access from a private browser window, then configure your apps and integrations with these credentials. Without password protection, any client device that can reach the HTTP service can access administrative and automation APIs.
Choose your language and region
The wizard launches automatically on fresh installations or after a factory reset. Select your interface language, then click the Next button to configure your location and regional preferences.
Your location coordinates are used for scheduled events like sunrise and sunset times. Select your preferred measurement system (Metric or Imperial). You can easily update these settings later from the administration menu.
Pick a starting configuration
On the next screen, choose one of the three setup modes:
Start with the default setup
Select this option for a guided setup of HomeGenie's core features. You can also include demo devices and enable optional modules for hardware, speech recognition, and voice synthesis.
Demo devices allow you to evaluate HomeGenie before configuring physical equipment. These simulated lights, switches, and sensors generate realistic activity so you can test how the dashboard behaves.
Enable only the interfaces you plan to use right away. Depending on what you pick, the wizard will prompt you for connection parameters or voice model selections. You can always click the Back button to adjust your choices before moving forward.
Configure manually
Choose manual configuration if you prefer to go directly to Settings and customize your system from scratch. This mode lets you install packages, activate interfaces, and adjust programs at your own pace. Consult the System settings guide to learn where each option is located.
Restore a saved configuration
If you have an existing backup, choose the restore option and upload your backup ZIP archive. HomeGenie displays a summary along with your user programs. Select the items you want to recover, then proceed to complete the restore process.
Make sure to use a system backup created with the Backup command under the Maintenance section of the settings menu. ZIP archives for software packages or custom widgets must be installed through their respective management tools instead.
Connect your hardware
HomeGenie communicates with physical equipment through dedicated interfaces. Enable the interface matching your hardware controller, then enter the parameters requested by the wizard. If you are starting with simulated devices only, you can set up physical connections later.
Smart-home controllers
The setup wizard includes built-in support for standard protocols such as ZigBee, Z-Wave, and X10. These interfaces connect HomeGenie to your hardware transceiver, which then communicates directly with your home devices.
For serial adapters, pick the assigned serial port and verify the communication parameters. The ZigBee configuration also prompts you for your hardware driver, offering choices like CC25XX, XBee, Ember, and ConBee. Keep in mind that adapters sharing the same protocol may still require different underlying drivers.
The video below demonstrates setup with ZigBee and Z-Wave hardware:
You can revisit these connection settings anytime in the Settings menu under Configuration > Interfaces. Other integrations, such as Philips Hue and KNX, are provided by add-on packages; see the Packages guide to explore available options.
GPIO, I2C, and SPI
On supported boards, GPIO / I2C / SPI enables programs that interact directly with hardware pins and buses. GPIO pins handle digital inputs and outputs, while I2C and SPI buses connect a wide variety of sensors and peripheral components.
For example, the GPIO Pin program can configure an output pin as a switch or an input pin as a sensor. Adjust its settings to match your wiring and connected components. Available examples and hardware capabilities depend on your board, operating system, and installed libraries.
This Raspberry Pi example configures an LED and runs a program to blink it:
You can use the built-in code editor to customize these examples in C#, JavaScript, or Python. You can also install hardware packages later without rerunning the setup wizard.
Configure speech recognition and synthesis
For voice interaction with an AI assistant, enable Local speech recognition for microphone input and Local speech synthesis for spoken replies. These two optional interfaces work independently and remain disabled by default.
- Speech recognition uses Whisper to transcribe microphone audio into text. Select your preferred model, language, and processing threads, or choose automatic language detection.
- Speech synthesis uses Piper to generate natural spoken responses. Select a voice model, then adjust speech rate and processing threads.
HomeGenie applies your settings automatically. If required voice models are missing, HomeGenie downloads them and tracks progress. You can pause, resume, or cancel downloads at any time. Disabling an interface stops its active download and hides its options.
An internet connection is required only for the initial model download. Once installed, speech processing runs entirely on the local HomeGenie computer. These voice models are completely separate from the language models used by Lailama.
HomeGenie stores voice models in the following directories:
- Speech recognition:
data/models/asr/ - Speech synthesis:
data/models/tts/
You can also configure speech recognition and synthesis later under Configuration > Interfaces in the Settings menu. If you prefer typing messages, you can safely skip downloading voice models during setup.
Finish setup and explore
When you proceed, HomeGenie installs your selected software packages and applies the configuration. Allow this process to complete before exploring the dashboard. Dashboards for chosen hardware interfaces are generated automatically once compatible modules are ready.
When the dashboard opens, explore a device widget to inspect its controls. If you enabled demo devices, use these virtual modules to familiarize yourself with the interface before connecting physical hardware.
Here are several recommended next steps:
- Settings — Configure system services or install additional software packages.
- Dashboards — Organize widgets around the devices you use most often.
- Scheduler — Create automated routines and timed events.
- AI Assistants — Connect Gemini Genie or Lailama for conversational home control.
Choose an AI assistant after setup
The AI chat becomes available as soon as initial setup is complete. You can open it from the dashboard and choose the assistant you want to interact with. Gemini Genie requires a cloud API key, while Lailama runs on a compatible local model with sufficient hardware resources. Follow the dedicated guide for each assistant to configure model settings and data permissions.
You can adjust your AI assistant and its capabilities directly in the chat window after setup. The speech options in the initial wizard configure voice input and output services, which work seamlessly with both local and cloud AI.
Use your voice in the chat
After configuring your assistant, open the chat options to enable local speech recognition or local speech synthesis. To dictate a message, start recording, click Stop when finished, review the transcription, and click Send to deliver your message. Spoken replies are played directly in your browser rather than through the server's audio output.
Grant your browser permission to access the microphone when prompted. To record voice messages from a smartphone, tablet, or another networked computer, you must access HomeGenie using a secure HTTPS address.
If your browser is running locally on the same computer hosting HomeGenie, microphone input is also supported over standard HTTP at this address:
http://localhost:PORT/Where PORT corresponds to the active port number used by HomeGenie.
You can always monitor the dashboard and type text messages over an ordinary HTTP connection to your server's local IP address.
Voice processing and AI reasoning operate independently. Even when speech recognition and synthesis run entirely on your local machine, your chat prompts are still sent to the cloud provider when using Gemini Genie.
Review the Local Audio guide for complete API examples and troubleshooting instructions.