
Google launched the Google Home Speaker on July 1, 2026 — its first new smart speaker in six years and its first device marketed as "built for Gemini." According to Google built a great smart speaker, but Gemini isn't ready for it, the $99.99 device delivers solid hardware but a frustratingly uneven AI experience. Reviewer Jennifer Pattison Tuohy spent five days testing it and came away impressed by the form factor, skeptical of the software.
The broader context here matters. Amazon revamped its Alexa hardware last fall. Google is now answering with hardware it hopes will anchor its smart home ecosystem around Gemini. The race to make AI assistants genuinely useful in the home is very much on — and this launch is Google's clearest statement of intent in years.
The Details: What Google Actually Shipped
At $99.99, the Google Home Speaker is a softball-sized sphere with a single full-range 58mm driver, three far-field microphones, a neural processing unit, Wi-Fi 6, Bluetooth 5.4, and Thread 1.3 border router support. It ships in four colors — jade, berry, hazel, and porcelain — with jade and berry exclusive to the US market, according to the review.
The hardware story is genuinely good. The Verge notes the design is clean, the light ring is subtle enough to disable in settings entirely (something no competing speaker offers), and the invisible touch controls are far more responsive than on previous Google speakers. It also pairs with the Google TV Streamer as an audio output device — a first for any Google Home audio speaker.
Audio quality lands in third place compared to the $99.99 Echo Dot Max and the $129 HomePod Mini, according to Tuohy's side-by-side testing. The reviewer notes bass is noticeably thin — a real-world consequence of the smaller single-driver design versus the Nest Audio's dual woofer-and-tweeter setup. It's loud and clear on mids and vocals, but it doesn't hit like its predecessor.
The Thread 1.3 border router capability is worth flagging for smart home power users. Google told the reviewer it is working on Thread 1.4 support, which is designed to improve interoperability between border routers from different manufacturers. That's a forward-looking detail that matters if you're building out a Matter-based home network.
Why It Matters: The AI Assistant Gap Is Still Very Real
Here's the honest read on what this launch tells us. Google shipped great hardware and an AI assistant that still isn't ready for prime time. That gap is the story.
The Verge's review documents some genuinely painful Gemini for Home failures. A complex multi-room command — turn off one room, adjust the thermostat, turn on lights in another — took Gemini 10 seconds to execute. Alexa Plus completed the same request in under three, per the review. Some cloud-dependent requests lagged close to 10 seconds. Even local commands like "Turn on the lights" occasionally hit the same lag.
Speed alone you could forgive. But the reliability issues are harder to wave away. The reviewer asked Gemini which voice options the speaker had — it repeatedly insisted there were none. There are. On a Nest Hub Max, the same query apparently triggered a hallucinated list of voice names including "Dimitrix" and "Russell Gethy." The actual voice options are named after plants. Gemini also announced the correct song title, then played an entirely different song. Twice.
The conversational upside is real, though. Tuohy notes Gemini handled natural language impressively — saying "I'm cooking and I don't want to get too hot" correctly triggered the AC to lower. That's the promise. But the reviewer also notes Gemini "has the memory of a goldfish" — it lost the thread of a recipe walkthrough after a brief pause mid-cook.
Gemini Live, which keeps the conversation open and remembers prior exchanges, is available — but it's locked behind a $10-a-month Google Home Premium subscription, and critically, it can't take actions for you. It only talks. For any action, it hands off to Gemini for Home. The reviewer describes it as feeling "bolted on."
If you work in SEO or digital marketing, you might be wondering what this has to do with you. The answer: a lot. Voice search and AI assistant queries are increasingly the front door to information — and if the assistants powering them are unreliable, that shapes which content formats and which sources actually get surfaced. An AI that confidently gives wrong answers and hallucinates product features is an AI that your clients' customers are going to stop trusting. That trust vacuum matters for how brands show up in AI-driven environments. AI visibility tracking is becoming a real discipline, not a nice-to-have.
What to Do Now
If you're an agency owner or in-house marketer, the Google Home Speaker story is less about buying a speaker and more about what it signals for the AI assistant landscape. Here's how to act on it.
- Audit how your clients' content performs in conversational AI queries. The Gemini-powered home assistant draws on the same underlying models shaping AI Overviews and other generative search experiences. If Gemini confidently gets basic facts wrong, it means the quality and structure of your content matters more than ever — not less.
- Don't over-index on one assistant platform. Amazon's Alexa, Google's Gemini, and Apple's Siri are all taking different paths. Build content and schema that serves structured, verifiable answers — not just one ecosystem's preferences.
- Watch the subscription paywall trend. Google Home Premium starts at $10 a month. Amazon's Alexa Plus carries its own subscription tier. Features being paywalled behind subscriptions changes how household AI assistants get used and how often — which affects query volume and behavior at scale.
- Test your structured data for home and local queries. Smart speakers are still heavily used for local lookups, business hours, and quick-answer queries. If your schema isn't tight, you're not competitive for the queries these devices field.
- Flag the hallucination problem for clients in regulated or reputation-sensitive industries. If an AI assistant confidently invents product features or wrong information about your client's brand, that's a reputation issue — not just a tech curiosity. It's worth building a monitoring practice around it.
Background and Context: Why This Launch Took Six Years
Google's last smart speaker before this one launched six years ago. That's a long time in consumer tech. In the meantime, the smart home category went through a rough patch — smart speaker sales plateaued, and Google killed or deprioritized several hardware lines. The Nest brand was folded into Google Home. Chromecast devices were consolidated. The category needed a reason to exist again.
Amazon moved first by pairing new Echo hardware with a revamped Alexa that the company positioned around large language model capabilities. Google's answer is the Home Speaker and Gemini for Home — but as The Verge's review makes clear, Google is shipping hardware that's ahead of the software built for it.
That's not a new story for Google. The gap between promising AI demos and dependable real-world performance has been a recurring theme across the company's launches over the past two years. Gemini for Home is another data point in that pattern. The conversational understanding is genuinely impressive in spots. The execution — speed, reliability, memory — still has real ground to cover.
One detail worth watching: the Verge review confirms Gemini for Home works on existing Google Home hardware — the Nest Audio, Nest Hub (2nd gen), and Nest Hub Max — not just the new speaker. That means Google isn't hard-locking AI features to new devices. Whether that makes the $99.99 upgrade worth it depends almost entirely on how fast Google improves the software.
If you want to get ahead of the AI visibility curve while it's still early — tracking where your brand and clients actually show up in generative AI responses — Aergos has tools built for exactly that kind of monitoring.
Frequently Asked Questions
Related Articles
Glossary terms in this article
Brush up on the definitions.
A type of AI model trained on massive text datasets to understand and generate human language at scale.
All marketing activities that use digital channels — search, social, email, display, content, and AI — to reach, engage, and convert target audiences.
A standardised format for providing information about a page and classifying its content so search engines can better understand it.
The extent to which a brand's content is referenced, cited, or surfaced in AI-generated answers from tools like ChatGPT, Gemini, and Perplexity.
AI systems that create original content—text, images, audio, or code—by learning patterns from training data.
When an AI language model confidently generates factually incorrect, fabricated, or unsupported information that sounds plausible but is false.

About Matt Weitzman
Senior SEO Strategist & Co-Founder
Matt has over 15 years of experience in technical SEO and digital marketing. He specializes in algorithmic recovery, enterprise architecture, and leveraging AI for content scaling. He is a frequent speaker at search marketing conferences.
More articles by Matt Weitzman

