A voice for your smart home
Ask your configured Home Assistant Assist pipeline for an action, and hear the response through the integrated speaker.
Voice DSP Plus Full Speaker
139,90€
Preorder: estimated shipping 10 October 2026. Not for immediate dispatch.
Your Home Assistant voice companion, already assembled. Four microphones used for beamforming and DOA, AEC for voice during playback, an ESPHome ESP32-S3 controller, a speaker with two passive radiators, a 360° LED ring and configurable gesture control.
Preorder: €139.90. Estimated shipping: 10 October 2026. Not for immediate dispatch.
Requires your own Home Assistant server and a configured Assist pipeline. Preloaded firmware; Wi-Fi and Home Assistant setup still required.
For Home Assistant
Your smart speaker.
Your smart speaker.
Your data.
Your choice.
A preassembled voice companion for Home Assistant. Four microphones, beamforming and acoustic echo cancellation help it hear your commands while the speaker plays.
You choose the voice services: a fully local Home Assistant pipeline, or the cloud services you prefer. Privacy depends on your speech recognition, conversation and speech synthesis settings.
Preassembled4 microphonesESPHome + Home Assistant
Preorder · €139.90. Estimated shipping: 10 October 2026. This is an estimated dispatch date, not a guaranteed delivery date. This product is not available for immediate dispatch.
Talk to your home. Listen to the reply.
Voice DSP Plus Full Speaker brings together RASPIAUDIO Voice DSP Plus and a CoreZero S3 ESP32-S3 controller running ESPHome in a compact, preassembled speaker. Four microphones, dedicated voice processing, powerful playback and touch-free controls, ready to connect to your Home Assistant setup.
Keep listening during playback
Dedicated microphone processing helps separate your voice from the speaker’s own sound, so you can speak without reaching for a button.
Make it your own
Configure gestures and interaction feedback in ESPHome, with your Home Assistant setup in control.
Voice, sound and touch-free control.
Your voice, even while music is playing
All four physical microphones are used by the audio software. They feed the microphone-array processing for beamforming and direction of arrival (DOA), rather than being reduced to a two-microphone input. Beamforming focuses on the voice; noise suppression reduces background noise; automatic gain control helps maintain a usable speech level.
Acoustic echo cancellation (AEC) uses the speaker playback as a reference to reduce that sound in the microphone signal. Ask for an action or stop playback while the speaker is playing, without reaching for a button. Results depend on playback level, placement and room acoustics.
Control your music with a gesture
The gesture sensor enables touch-free interactions, configurable in ESPHome to suit your preferences. For example, you can assign left/right gestures to volume, an upward gesture to the next track, and another supported gesture to the previous track. These are configuration examples, not a guarantee of a particular factory gesture mapping; track controls also depend on the connected media source.
A 360° LED ring provides visual feedback for interaction and sound direction.
A compact speaker, built to evolve
- Speaker with two passive radiators: an enclosure designed to support bass reproduction, for spoken replies and music.
- 3D-printed enclosure: supplied assembled, with the microphones, electronics and speaker already integrated.
- 3.5 mm stereo output: connect headphones or a compatible external powered audio system. Start at a low output level when feeding an amplifier.
- Internal I2C expansion: a route for compatible accessories and custom integrations, subject to electrical and firmware compatibility.
- External Wi-Fi antenna: included with the ESP32-S3 controller.
Inside the system
| ESP32-S3: | dual-core 32-bit LX7, up to 240 MHz, 512 KB internal SRAM, 16 MB flash and 8 MB PSRAM. |
|---|---|
| Audio processor: | XMOS XVF3800, dedicated to microphone-array voice processing. |
| Audio functions: | four-microphone beamforming, AEC, noise suppression, AGC and direction of arrival. |
| Application platform: | ESPHome, connected to your Home Assistant installation. |
Everything in the pack
Two RASPIAUDIO boards, one assembled speaker. Both boards below are included, along with the enclosure and accessories.
- RASPIAUDIO Voice DSP Plus board with four microphones.
- CoreZero S3 ESP32-S3 controller, running ESPHome.
- 3D-printed enclosure.
- Speaker and two passive radiators.
- 360° LED ring.
- Gesture-control sensor.
- External Wi-Fi antenna.
Connect it to your Home Assistant
The system is supplied with a preloaded ESPHome profile: no assembly or initial firmware flashing is required for normal setup. Connect it to your Wi-Fi, add the device through Home Assistant’s ESPHome integration, and choose your Assist pipeline.
You need an existing Home Assistant server and a configured Assist pipeline. Neither a Home Assistant server nor a Raspberry Pi is included. The ESP32-S3 runs the local wake-word component; the XVF3800 handles the microphone audio front end. Speech recognition, answers and speech synthesis depend on your Home Assistant pipeline, which may use local or cloud services.
The controller is listed as ZeroCore S3 in the RASPIAUDIO app installer. For recovery or manual firmware installation, use the ZeroCore S3 / Home Assistant browser installer and its setup notes. ESPHome OTA is also available for custom builds. Updates are user-initiated; the product does not silently install every new published release.
Home Assistant firmware & setup
Use the controller’s USB port for recovery or manual installation. Follow the installation notes for the assembled ESPHome system.
Voice DSP Plus board resources
The dedicated audio board has its own firmware and documentation. Its USB loader is for standalone USB audio modes, not for installing the Full Speaker Home Assistant application.
