The smart home landscape is in the midst of an important transition. After years of relying almost entirely on cloud-based voice assistants, a new wave of locally processed voice hardware is emerging. These devices promise greater privacy, reduced latency, and independence from commercial ecosystems. Two of the most talked-about entrants in this space are the Home Assistant Voice Preview Edition and the Satellite 1 Developer Kit from Future Proof Homes. Both are designed to run voice commands locally with Home Assistant, giving homeowners the ability to manage their smart home without relying on cloud servers. Yet the performance of early private voice hardware remains a topic of debate, especially when compared with established cloud devices like the Amazon Echo Dot.
Based entirely on extensive real-world testing, this article explores how Home Assistant Voice PE and Satellite 1 compare to each other and to the Echo Dot. The goal is to help readers understand where private voice hardware stands today, where it still struggles, and what the future likely holds for these open, privacy-focused devices.
The New Generation of Local Voice Assistants
The emergence of local voice assistants stems from a clear need among smart home enthusiasts for privacy, flexibility, and autonomy. The Home Assistant Voice Preview Edition, released in December 2024, represents one of the first consumer-facing attempts to bring local speech processing into homes without depending on major cloud platforms. Future Proof Homes soon followed with the Satellite 1 Developer Kit, aimed at users who prefer a more modular and experimental approach. Although both devices promise privacy, offline processing, and compatibility with Home Assistant, they target slightly different audiences. Understanding their hardware philosophy helps illustrate why their performance differs.
Unlike traditional smart speakers, these devices are deeply tied to open-source tools like ESPHome and Home Assistant’s Assist voice system. They operate without requiring an internet connection and rely on community-driven software pipelines. This shift represents a meaningful step toward decentralized voice control and raises expectations among early adopters for device speed, reliability, and ease of use.
Hardware Form Factors and Build Differences
Each device offers a distinct hardware approach that reflects its intended purpose. Home Assistant’s Voice PE and the Amazon Echo Dot arrive fully assembled, with the core electronics housed inside a polished enclosure meant to blend into a home environment. Voice PE features a compact square shape, whereas the Echo Dot maintains its familiar round design. Both occupy similar space on a desk or shelf, offering a plug-and-play experience.
Satellite 1 takes a more experimental route. It ships as a kit consisting of two separate boards. A circular hat board houses four microphones, four programmable buttons, a headphone jack, a built-in amplifier, and environmental sensors. A rectangular core board, built around an ESP32 S3, connects directly into the hat using a 40-pin connector. While an enclosure was not available at the time of purchase, this open design appeals to developers and hobbyists who prefer the freedom to extend, mount, or reconfigure the hardware.
Microphone arrays also differ across devices. Voice PE includes two microphones, while the Satellite 1 and the Echo Dot both use four. Satellite 1, however, requires a firmware upgrade to enable all microphones. Both Voice PE and Satellite 1 rely on XMOS chips for audio processing and far-field voice detection. The Echo Dot uses a MediaTek MT516 platform. These differences explain why capture sensitivity and response quality can vary, especially in room environments with ambient noise.
Speakers, Audio Output, and Physical Interaction
Voice-based devices typically double as smart speakers, but the three devices take diverging paths in speaker design. Voice PE and the Echo Dot include built-in speakers rated at 3 watts and 1.2 watts respectively. Both also provide 3.5 mm output for external audio. Satellite 1 stands out because it does not include an integrated speaker. Audio must be heard through headphones or an external speaker, although the output can drive speakers up to 25 watts, which is significantly more powerful than the other two. This gives Satellite 1 an advantage for users who wish to integrate it into a larger or custom audio environment.
All three devices include physical controls, though implementation varies. Voice PE provides an action button, a mute switch, and a tactile rotary dial for volume control. Satellite 1 and the Echo Dot offer four buttons for mute, actions, and volume adjustment. Satellite 1’s buttons are fully programmable, which opens up possibilities for custom automations or integrations. These differences reflect a broader divide between those seeking an appliance-like voice assistant and those seeking a flexible platform.
Lighting indicators remain an important part of the user experience. Voice PE uses twelve LEDs, and the Satellite 1 uses twenty-four. Both can be manipulated through Home Assistant automations, making them useful for visual notifications. The Echo Dot employs its well-known color ring with eight built-in animation states. Cloud-based devices like the Echo Dot provide more predefined lighting behaviors, while the open-source alternatives rely on customization.
Sensors, Expansions, and Additional Features
Smart home automation often extends far beyond voice commands, and the devices differ in how they support sensors and accessories. Voice PE reserves a growth port for future accessories, although expansions must be purchased separately. Satellite 1 includes temperature, humidity, and luminosity sensing out of the box. It also supports optional millimeter-wave presence sensors, such as the LD2410 for still detection and LD2450 for multizone detection. The included JST cable allows flexible sensor placement, which is ideal for advanced automation scenarios.
The Echo Dot, in contrast, tightly integrates with Amazon’s ecosystem. Its biggest advantage lies in its library of Alexa skills, numbering in the tens of thousands. These enable timers, games, routines, news, music playback, and a wide range of third-party services. This reflects a major distinction between cloud-first ecosystems and open-source platforms. Cloud systems excel in breadth, while local systems emphasize privacy and control.
Power requirements also differ. Both Voice PE and Satellite 1 use USB-C for power, but Satellite 1 demands USB-C Power Delivery to support high-wattage external speakers. Neither includes a power adapter. The Echo Dot includes a power cable, but it uses a proprietary barrel connector. These small details can affect usability, especially for mainstream users.
Software Ecosystems and Connectivity
One of the most significant differences among the devices lies in software. Home Assistant Voice PE and Satellite 1 run on ESPHome firmware with native integration into Home Assistant. They function entirely offline and depend on locally configured voice pipelines. When required, they fall back to a large language model agent. In the test environment, both Voice PE and Satellite 1 used a local conversation agent with fallback to Google Gemini for broader queries. Because they shared the same configuration, they produced identical spoken responses. What varied was processing time and reliability.
The Echo Dot runs Alexa device software and relies heavily on cloud services. Internet connectivity is mandatory for full functionality. Its integrations and natural language understanding are built around Amazon’s broader ecosystem, which benefits from massive datasets, optimized machine learning pipelines, and dedicated cloud processing. This creates a speed and accuracy advantage that local hardware struggles to match today.
All three devices support Wi-Fi and Bluetooth. The Echo Dot also functions as a Matter hub, which Homes Assistant devices do not attempt to replicate. Cloud assistants typically try to unify as many devices as possible under a simple interface, while Home Assistant devices emphasize customization.
A Controlled Test: How Each Device Responded
To evaluate performance, each device was placed in the same location, connected to the same Wi-Fi network, and given access to the same smart home entities. Ten common voice commands were issued to assess speed, understanding, and reliability. These commands covered household tasks, timers, lighting control, temperature checks, conversational queries, and music playback.
Before beginning the test, all devices were configured to ensure fairness. Voice PE and Satellite 1 used identical voice pipelines. The Echo Dot used Amazon Music for playback, while the other two used Music Assistant within Home Assistant. All devices were measured by timing the interval between the user’s final spoken word and the start of the device’s audible response.
The test revealed a substantial performance gap. The Echo Dot consistently responded in about two seconds. Voice PE averaged around five seconds, and Satellite 1 averaged about six seconds. In an early run of the test, Satellite 1 took roughly twenty-one seconds to respond, but after switching to a different text-to-speech voice within the Home Assistant pipeline, its response times aligned more closely with Voice PE. This suggests that certain voice models produce audio at different bit rates, affecting decoding time and overall latency.
However, response time was only part of the story. The Echo Dot successfully completed all ten tasks, while Voice PE and Satellite 1 succeeded in only five. This demonstrated that reliability, not just speed, remains a challenge for early local voice systems.
Where Local Voice Assistants Struggled
Experts highlight several areas where Voice PE and Satellite 1 fell short. These issues affected daily usability and contributed to frustration.
Temperature Queries and Location Awareness
Neither device could answer the question about the room temperature, even though all were assigned to the same location with the temperature entity exposed. When told “I’m cold”, the devices requested additional clarification about the user’s location rather than inferring it from the device context. This contrasts sharply with the Echo Dot, which understands that a room-specific request should reference its assigned location.
Food and General Knowledge Queries
When asked for recipe suggestions using specific ingredients, both Voice PE and Satellite 1 read aloud the word “asterisk” repeatedly. This occurred because their text-to-speech engine interpreted formatting characters literally. The Echo Dot delivered a recipe recommendation smoothly.
Clarifying Questions
When Voice PE and Satellite 1 required follow-up answers, they did not continue listening automatically. Instead, they required the user to repeat the wake word before responding. This created confusion and disrupted the natural flow of conversation. The Echo Dot handled follow-ups as expected.
Timer Inconsistencies
Timers exhibited inconsistent behavior on Voice PE, such as requiring a timer name some days but not others, and struggling to keep track of multiple active timers. These issues did not appear on the Echo Dot, which managed timers reliably.
Music Playback
Voice PE and Satellite 1 were unable to play music from the local library, even when specifying the correct media player. This appears to be a limitation of Music Assistant, which may not yet support voice control. The Echo Dot handled music playback without any trouble.
Why Cloud Assistants Excel Today
The test results reinforce a broader pattern. Cloud-based devices like the Echo Dot benefit from massive training datasets, optimized infrastructure, and years of engineering refinements across latency, natural language processing, device context, and fallback logic. Their conversational flow is more polished, and their handling of edge cases is more robust. Although local devices prioritize privacy and control, they currently lack the same level of maturity.
Experts also notes that the overall experience with Voice PE was frustrating enough that the household eventually switched back to using the Echo Dot in the kitchen after several weeks. For mainstream users, reliability trumps philosophical advantages. Until private voice hardware becomes more consistent and intuitive out of the box, this will remain a challenge.
The Path Forward for Private Voice Hardware
Despite the shortcomings observed in the test, the future for private voice assistants remains promising. The Home Assistant Voice PE and Satellite 1 are clearly labeled as preview and developer products. Their purpose is to give enthusiasts and developers early access to local voice technology so that the community can refine and improve it. Early adopters expect a degree of experimentation, and the open-source ecosystem tends to evolve rapidly with community feedback.
There are also variables that could influence performance. Some users report excellent results when running Home Assistant Cloud, different hardware combinations, or more powerful servers. Additional tuning of entities, aliases, and LLM instructions can also improve response rates. Weather entities can be customized to support better forecasting queries. These experiments show that the ecosystem is flexible, but flexibility alone does not guarantee a mainstream-ready experience.
As development progresses, private voice systems will likely benefit from advances in embedded speech recognition, improved text-to-speech pipelines, more refined fallback logic, and broader support for local media playback. Their biggest long-term advantage remains privacy. Users who prioritize data control will continue to support these efforts, accelerating improvements.
Conclusion
Local voice assistants represent an important evolution in smart home technology. Devices like the Home Assistant Voice Preview Edition and the Satellite 1 Developer Kit show that offline voice control is becoming more accessible, customizable, and community driven. Yet the testing demonstrates that they still lag behind mature cloud-based solutions in speed, reliability, natural conversation flow, and task completion.
Based on the evaluation, the Amazon Echo Dot outperformed the two private devices in every metric. It was faster, more accurate, and more consistent. While this outcome was expected given the Echo Dot’s years of development, it also shows that private voice hardware remains in an early stage. Early adopters must be willing to tolerate inconsistencies while contributing to the ecosystem’s growth.
The trajectory for local voice control is encouraging. The technology will improve, the pipelines will become more refined, and hardware maturity will continue to grow. For now, however, mainstream users may prefer the reliability of cloud devices, while privacy-focused users will welcome the progress and potential of these open-source alternatives.
