Google officially began pushing Android 17 to Pixel devices on June 16, but Gemini Intelligence, an agent-style AI suite that Google positions as a core selling point, will only start rolling out in batches to the Google Pixel 11 series and Samsung’s upcoming flagship models this summer, with broader availability coming later this year.
Google June 16 Pushes Android 17: Three Initial Wave Features
The Android 17 initial wave update includes three directions:
Multitasking interface (Bubbles): A new feature lets users shrink individual apps into bubble form, freely stack them, and drag them to any position on the screen. The folded-mode taskbar adds a dedicated Bubbles section, optimized for large-screen multitasking.
Community video (Screen Reactions): The system layer natively includes a “screen reactions” feature. While recording, users can overlay their face on top of what the phone is currently displaying—creating a “camera + screen in frame” format.
Security upgrade: Adds “one-time location access,” ending an all-or-nothing location authorization mode. In sparsely populated regions, “approximate location” intelligently expands the display range. Lost mode is strengthened to biometric identification plus PIN-code two-factor authentication.
Gemini Intelligence Hardware Requirements: Gemini Nano v3 + 12GB RAM, Most Older Devices Excluded
Gemini Intelligence is the agent-style AI suite for Android 17. It can act on the user’s behalf to complete multi-step chained tasks on the phone, including organizing a calendar, summarizing webpages in Chrome, creating custom widgets using natural language, and using the tool “Rambler” to polish voice messages.
The reason for the delay is hardware requirements, not software progress. Gemini Intelligence requires the device to be equipped with Gemini Nano v3 (a small on-device language model running directly on the phone’s chip), a qualified flagship SoC, and at least 12GB RAM. These requirements will exclude most Android phones released before 2026, and even some mid-to-high-end models are no exception.
Gemini Intelligence to Roll Out in Batches This Summer to Pixel 11 and Samsung Flagships; Wider Release Later This Year
The rollout schedule for Gemini Intelligence is: this summer, it will first open in batches to the Google Pixel 11 series and Samsung’s upcoming flagship models, with broader availability expected in late 2026. The Android 17 version Google is currently pushing includes a complete system infrastructure upgrade, but does not include Gemini Intelligence functionality; the two belong to different update waves.
Common Questions
Which Pixel devices can upgrade to Android 17 immediately?
According to Google’s announcement, there are 21 Pixel devices being pushed from today, covering Pixel 6, Pixel 6 Pro, Pixel 6a, the Pixel 7 series, the Pixel 8 series, the Pixel 9 series (including Pro Fold), the Pixel 10 series (including Pro Fold and 10a), Pixel Tablet, and Pixel Fold. Updates for other Android brands will follow throughout 2026.
Which phones are excluded by the Gemini Nano v3 and 12GB RAM thresholds?
According to reports, Gemini Intelligence requires the Gemini Nano v3 on-device model, a qualified flagship SoC, and at least 12GB RAM. This means most Android phones released before 2026, including some mid-to-high-end models, can’t run Gemini Intelligence due to insufficient RAM or an outdated SoC generation, though they can still upgrade to Android 17’s other functions.
How do Apple and Google’s AI update plans differ and what’s the same?
Reports say Apple announced last week its AI and Siri improvement plan launching this fall, while Google rolled out Android 17 in the same period and announced that Gemini Intelligence would open up this summer. A common feature of both is that the truly agent-style AI functions have not yet officially reached users; both are currently in a stage of “preview direction and managing expectations.”
