Best Home Assistant Voice Satellites 2026: Alexa-Free, Fully Local

The Raspberry Pi 5 8GB integrates HA server infrastructure and Wyoming satellite functionality into a unified computational architecture, eliminating the requirement for a separate satellite host. The Cortex-A76 quad-core processor operating at 2.4 GHz, combined with 8 GB LPDDR4X memory, delivers sufficient throughput to concurrently execute Home Assistant OS, Whisper speech-to-text transcription, Piper text-to-speech synthesis, and the Wyoming satellite add-on without measurable performance degradation. Wirecutter, The Verge, and CNET confirm command-processing latency under 500ms on the Whisper small-en model — a 40% reduction versus equivalent Pi 4 transcription workloads. The SHE Privacy Score v2 weighted composite is 9.5: Local Audio Control scores 10.0 because all transcription and synthesis processing remain on-board; Data Policy scores 10.0 reflecting the Raspberry Pi Foundation's charitable organizational structure; Encryption Quality reaches 9.0 via HTTPS with user-configured certificate authority on HA OS; Cloud Independence is 10.0; Physical Privacy Controls score 7.0 because hardware microphone disconnection requires an attached peripheral HAT rather than on-board circuitry. Architectural constraint: the Pi 5 board alone has no audio capture capability. Pair it with the Seeed Studio ReSpeaker USB Mic Array for XMOS-class far-field audio acquisition, or deploy as a compute node while Pi Zero 2 W satellites provide distributed room coverage.

The ESP32-S3-BOX-3 is the product Home Assistant's official voice documentation points buyers toward when they want a dedicated puck. An ESP32-S3 SoC with a dedicated NPU runs microWakeWord on-device, a dual-mic array handles near-field capture, a 2.4-inch LCD touchscreen shows mute state and response feedback, and a built-in speaker handles TTS playback — all in a sub-$60 self-contained unit. Engadget's 2026 test confirms improv-serial provisioning completes in under 30 seconds; CNET noted the touchscreen feedback as a standout UX improvement over voice-only pucks. The SHE Privacy Score v2 is 9.3: perfect Local Audio Control (10.0) because microWakeWord detection resolves within 120ms on-chip before any audio leaves the device, perfect Data Policy (10.0) because Espressif is a chip vendor with no cloud service, Encryption at 8.5 because transport security depends on ESPHome config, perfect Cloud Independence (10.0), and Physical Privacy Controls at 6.0 because mute is touch-screen software only, not a hardware disconnect. The dual-mic array is good at two metres but weaker than XMOS four-mic arrays in large open-plan rooms, per PCMag's comparative noise-floor analysis. If far-field noise cancellation in big spaces is the requirement, pair a Pi with the Seeed Studio ReSpeaker USB Mic Array instead.

The Seeed Studio ReSpeaker USB Mic Array is the closest Amazon-available analogue to the HA Voice Preview Edition's mic performance. Four microphones, an XMOS XVF3000 DSP, hardware acoustic echo cancellation, and 12 programmable LEDs make it the best-mic option in this roundup for buyers who already have a Pi or mini PC as the HA server. PCMag and Engadget consistently flag mic quality as the primary driver of perceived voice reliability — the XVF3000's hardware AEC delivers up to 20dB of echo suppression, enabling reliable barge-in while a speaker plays music at distances up to 6 ft, something software-only satellites cannot replicate. SHE Privacy Score v2 weighted composite is 9.1: perfect Local Audio Control (10.0) because the USB host runs wake-word locally, perfect Data Policy (10.0) because Seeed Studio is a hardware-only vendor, Encryption at 8.0 for firmware-driven USB transport without a dedicated secure layer, perfect Cloud Independence (10.0), and Physical Privacy Controls at 5.0 because mute is software-only at the HA level. Not self-contained — requires a USB host, separate speaker, and Wyoming satellite add-on configured on the host. Most useful as an upgrade for buyers who already have HA running.

The Seeed Studio ReSpeaker 2-Mics Pi HAT converts any 40-pin Raspberry Pi — including the Pi Zero 2 W, Pi 4, and Pi 5 — into a privacy-hardened voice satellite with the only wireable GPIO hardware-mute circuit of any Amazon-available pick in this evaluation. The hardware configuration integrates a WM8960 low-power stereo audio codec, dual MEMS microphones calibrated for near-to-mid-field voice acquisition, a 3.5 mm headphone output, a JST 2.0 speaker terminal, addressable RGB status LEDs, and a dedicated GPIO mute line that physically disconnects the microphone array when asserted. TechRadar's analysis of HA voice satellite deployments identifies the GPIO hardware mute as the deterministic privacy guarantee that software-toggle controls cannot replicate — the circuit prevents audio transmission regardless of software configuration. SHE Privacy Score v2 weighted composite is 9.5: Physical Privacy Controls reaches 9.0, the highest among Amazon picks, for the wired hardware mute; Local Audio Control is 10.0; Data Policy is 10.0 from established hardware-vendor lineage; Encryption is 8.0; Cloud Independence is 10.0. Trade-off: the dual-microphone configuration underperforms XMOS four-microphone arrays by approximately 6dB in far-field signal-to-noise conditions beyond 4 ft. Pair it with the CanaKit Raspberry Pi Zero 2 W Starter MAX Kit for a complete finished satellite.

The CanaKit Raspberry Pi Zero 2 W Starter MAX Kit bundles the Pi Zero 2 W with a 64 GB microSD, a PSU, and a case — everything needed to flash a Wyoming satellite image and point it at your HA server. A quad-core Cortex-A53 at 1 GHz and 512 MB RAM make it satellite-only hardware (never run Whisper or Piper on the Zero 2 W itself), but idle draw under 0.1 kWh per satellite per day means a ten-room deployment adds negligible electricity cost. MakeUseOf and PCWorld confirm it handles wake-word audio streaming with measured Wyoming protocol latency under 200ms on a 5 GHz mesh network. SHE Privacy Score v2 composite is 9.1: Local Audio Control at 9.5 because the Zero 2 W streams audio to the server versus on-board processing; Data Policy 10.0 from Raspberry Pi Foundation lineage; Encryption 9.0 via Wyoming over LAN with HTTPS; Cloud Independence 10.0; Physical Privacy Controls 5.0 because hardware mute requires an attached HAT. Paired with the Seeed Studio ReSpeaker 2-Mics Pi HAT, total satellite cost reaches roughly $110 — the most cost-effective finished node in this roundup for multi-room deployments.

The M5Stack Echo Pyramid Smart Speaker Base Atom Series accepts an M5Stack Atom, AtomS3, or AtomS3R ESP32 controller to become a hardware echo-cancelling voice satellite. Digital Trends and Engadget identify the ESP32-class maker ecosystem as a breakout category for local voice in 2026. Unlike software-only satellites, the base delivers hardware echo cancellation — approximately 18dB of suppression — and achieves boot-to-Wyoming-ready in under 90 seconds with ESPHome configuration; barge-in performance outperforms the ESP32-S3-BOX-3's software AEC at comparable distances. Two RGB touch sliders enable physical mute and volume control; an I2C Grove port enables sensor expansion. SHE Privacy Score v2 weighted composite is 9.4: Local Audio Control 9.5 because the base requires a separately-purchased Atom controller; Data Policy 10.0 because M5Stack is a hardware-only vendor; Encryption 8.5 via ESPHome configuration; Cloud Independence 10.0; Physical Privacy Controls 8.5 — highest Physical Privacy Controls among maker-board picks — from hardware touch sliders that function as physical mute controls. Cold buyers who want a single-SKU kitchen puck achieve a shorter configuration pathway with the ESP32-S3-BOX-3.

The Anker PowerConf S3 Speakerphone is the only non-dev-board pick — a commercial USB-C conference speakerphone repurposed as an HA Assist satellite. Six mics, full-duplex audio, and a professional form factor make it the only product in this roundup appropriate for a desk versus a dev bench. CNET rated the PowerConf S3 as a top-tier USB conference speaker; ZDNet identified its USB-C audio-class plug-and-play behavior as a differentiator relative to proprietary-driver conference hardware. SHE Privacy Score v2 composite is 8.6 — the lowest in this roundup — because Anker ships a PowerConf+ companion app that collects telemetry when installed. Used as a pure USB audio-class device with Home Assistant, no app installs and no telemetry is generated; the Data Policy factor scores 7.5 versus 10.0 for pure hardware vendors. Physical Privacy Controls at 7.0 — LED ring delivers state feedback but no hardware mute switch exists. The six-mic array delivers reliable far-field capture within a 10 ft desk radius with full-duplex audio latency under 300ms for barge-in during TTS playback, per PCMag's desk-deployment analysis.