Google reinvents the smart speaker with Gemini in June Pixel Drop

Google is rebuilding the smart home speaker around Gemini, adding natural-conversation capabilities — mid-sentence corrections, two-way back-and-forth dialogue, and 10 new voices — that aim to make the device feel like a real assistant rather than a command parser. The push lands as the assistant battleground heats up, with Apple's rebuilt Siri AI (itself reportedly Gemini-powered) impressing reviewers and expanding to Apple Watch the same week.

The broader June 2026 Pixel Drop and Android 17 release expand Gemini across Google's surface: Gemini Omni video editing for Pro subscribers, the Lyria 3 music generator, and conversational photo editing extended to Pixel 6 and newer across more European markets. Android 17 also shipped new multitasking tools. Together these reinforce Google's strategy of embedding Gemini everywhere rather than relying on a single hero product.

The competitive context is sharp and ironic. Google is simultaneously powering Apple's Siri (a Gemini deal) and competing with Apple on devices, while losing Gemini co-lead Noam Shazeer to OpenAI in the same week — a reminder that Google's model leadership is under both talent and product pressure. Jeff Dean separately highlighted Google's TPU training-supercomputer lineage from TPU v2 to Ironwood, underscoring its infrastructure depth.

The caveat: Gemini 3 Deep Think drew mixed reviews this week, with developers warning of a 'hallucination trap' on long reasoning and pricing gripes about a '$250 extra' tier. Conversational speaker demos are easy to stage and hard to ship reliably. Watch whether the new voices and two-way conversation hold up against real household noise and interruptions.