OpenNextion ESPHome Voice Assistant is a flash‑and‑use hardware solution that turns your OpenNextion smart display into a voice control panel for your smart home. Just connect a microphone, speaker, and WiFi, then integrate with Home Assistant to instantly get a voice terminal with local wake‑word detection, touch control, and real‑time status feedback.
| Feature | Description |
|---|---|
| Local Voice Wake‑Up | Built‑in "Okay Nabu" wake word by default – detects locally without an internet connection for fast response and privacy protection. |
| Full‑Duplex Voice Interaction | Ask questions via the microphone; the device connects to Home Assistant for speech recognition and processing, then plays voice replies through the speaker. |
| Touch‑Based Status UI | The screen clearly shows current states: Listening, Thinking, Replying, Error, Mute, or Timer Ringing. |
| Timer Ringing & Touch Stop | Set a timer with a voice command such as “set a timer for 5 minutes”; the device will ring when the timer ends, and you can stop it by tapping the screen. |
| WiFi Provisioning | On first use, the device automatically starts a provisioning hotspot – easily configure WiFi via your phone or computer browser. |
| Home Assistant Auto‑Discovery | After provisioning, the device is automatically discovered by Home Assistant and added as an ESPHome device, seamlessly integrating into your smart home system. |
| Simple Initialization | Press and hold the BOOT button on the device for 10 seconds to clear previous settings and enter the initial setup state. |
Note: The ESP32 device provides microphone and speaker interfaces, but these modules are not included and must be obtained separately. You may refer to officially compatible accessories or use other compatible hardware.
Before starting, ensure the USB cable supports data transfer and maintain a stable connection between the board and your computer.
After successful flashing, the board will automatically restart and boot into the ESPHome voice assistant.
Before building a local smart home voice assistant, you need to download local Speech‑to‑Text (STT) and Text‑to‑Speech (TTS) model apps from the Home Assistant Add‑on Store. The official recommended apps are:
STT - Whisper
TTS - Piper
After installation and enabling, go to Settings > Devices & Services. Home Assistant will automatically discover the newly installed apps. Click Add for each to add them. Once added, they will appear under the Wyoming Protocol integration – your Home Assistant voice assistant now has local STT and TTS capabilities.
Go to Settings > Voice Assistants to add or modify a voice assistant. Configure its name, conversation agent, language, and select the STT/TTS models. Choose the Whisper model for Speech‑to‑Text and the Piper model for Text‑to‑Speech to complete your local voice assistant setup.
*Make sure the device and Home Assistant are on the same local network.
onx2432g028-xxxxxx (open)http://192.168.4.1
*Make sure to select "Full local processing" for voice communication to build a local voice assistant.
*Be sure to select the Voice Assistant you have just configured.
Now your voice assistant is ready. The final – and most important – step is to give it access to the entities you want to control, so it understands your commands.
Go to Settings > Devices & Services > ESPHome Integration, click the configuration icon for your ESP32 device, and select “Allow the device to perform Home Assistant actions” – this enables the ESP32 device to control entities in Home Assistant.
Now, after completing the above configuration, you can start a conversation with your new local voice assistant – use voice commands to effortlessly control switches, outlets, lights, and other connected devices!
💥To reconfigure WiFi, press and hold the BOOT button on the device for 10 seconds to initialize, or click the "flashing settings" icon in the top‑right corner and check "Erase all flash before flashing" to factory reset the ESP32 device, then re‑flash the firmware.
Yes. This project essentially turns the OpenNextion smart screen into a voice input/output terminal for Home Assistant. It must work with Home Assistant to complete full functions such as speech recognition, dialogue processing, and speech synthesis.
No. The OpenNextion screen itself only provides the interfaces for microphone and speaker. You need to purchase the microphone (Nextion Dual MIC Board) and speaker (Nextion BOX Speaker) accessories separately.
The default wake word is "Okay Nabu", which is built into the project. If you want to customize the wake word, you need certain technical skills to rebuild the firmware from source code and configure it accordingly.
Latency mainly depends on the performance of your Home Assistant server, network conditions, and the STT/TTS pipelines used. Under good local network and smooth server conditions, the experience will be quite fluid. The project documentation also notes that weak Wi‑Fi signals may cause TTS audio playback to stutter.
The screen intuitively displays Listening, Thinking, Replying, Error, Mute, and Timer ringing statuses, so you always know the current working stage of the voice assistant.
We have designed a matching 3D‑printed enclosure, which you can obtain from the Resources section. If you don't have a 3D printer, you can use a third‑party printing service, or simply use the board bare. The enclosure will reserve routing space for the microphone and speaker.