Gemini for Home Adds Continued Conversation Feature
TL;DR
After your first 'Hey Google' command, the microphone stays active briefly: pulsing lights on Nest devices signal it is listening: so follow-up questions land without repeating the wake word. The update ships over-the-air to existing Nest Hub, Nest Mini, and Nest Audio devices.
What changed
Google rolled out Continued Conversation for Gemini for Home on April 21, 2026. This feature activates after your initial "Hey Google" command: the microphone stays open briefly, shown by pulsing lights on Nest devices, letting you follow up without repeating the wake word.
It brings four upgrades over prior versions. Gemini now holds conversational context across turns, unlike the old Google Assistant. Support covers all global languages and regions. AI detects side-talk better to avoid false triggers, and the feature works household-wide once enabled.
Why it matters
Continued Conversation cuts friction in voice AI, making Gemini feel more like talking to a person than issuing commands. Google bets on this to boost daily adoption on Nest hardware, gathering more user data from millions in early access to refine models.
It pressures Amazon Alexa and Apple Siri, which lag in context retention. For Vibe Builders automating home offices, this means reliable voice triggers for Notion updates or Zapier flows via smart speakers, without constant wake words interrupting workflows.
How to use it
You need a Nest speaker or display with Gemini for Home enabled, available in supported regions. Open the Google Home app, go to Home Settings, select Gemini for Home voice assistant, then toggle Continued Conversation on.
Say "Hey Google" to start, watch for pulsing lights, and follow up naturally. Test multilingual if needed; it rolls out globally now, no U.S. English limit.
Watch for
Success shows in user retention metrics or expanded context windows beyond seconds. It breaks if side-talk false positives rise in noisy homes. Expect next: deeper integrations with Google Home routines or third-party APIs for builder automations.
Harsh’s take
Gemini for Home's Continued Conversation fixes a basic pain point in voice AI: no more wake-word spam mid-chat. But for a solo Vibe Builder in 2026, this is table stakes, not a workflow booster unless you pipe it into Zapier via IFTTT or custom routines. It shines in home-office setups where you dictate Notion pages hands-free, yet the short listen window and household-wide access risk privacy leaks from guests barking orders.
The real trade-off hits data hawks like Google: millions of opted-in users feed model training, but your convos become training fodder too. Builders get smoother voice inputs for quick automations, like querying ChatGPT summaries aloud, only if Nest hardware sits in your stack.
Do this now: if you own a Nest speaker, enable it today and test chaining three follow-ups into a Zapier task. Skip if you're all-in on Alexa; the context edge isn't worth switching ecosystems yet.
by Harsh Desai
About Gemini
View the full Gemini page →All Gemini updatesGo deeper
More AI news
- FeatureDependabot adds three-day cooldown before version updates
Dependabot now defaults to a three-day cooldown that delays version update pull requests.
- FeatureHermes-agent enables smart approvals by default and adds user-defined deny rules
Hermes-agent now defaults to smart LLM approvals for flagged commands and evaluates each independently. It adds user-defined deny rules that block commands even in yolo mode with explanatory feedback via /deny.
- FeatureHermes-agent reduces first-token latency by 80% across platforms
Hermes-agent cuts cold-start initialization from 4.3s to 0.9s across CLI, gateway, TUI, and desktop. Live reasoning streams by default with per-token response rendering.