Google Gemini Adds Video, Music, Projects on Android
Google is rolling out a significant update to Gemini on Android through Google app version 17.32, according to NokiaPowerUser. The update expands the assistant's Add menu with new input types: video clips, music files, projects, and guided learning sessions. NokiaPowerUser also reports that many Gemini features can now be accessed from a floating overlay without launching the full Gemini app, letting users stay within their current workflow. The update aligns with Google's broader strategy of making Gemini a persistent, system-level assistant on Android rather than a standalone chat app. For platform and ML engineers, adding richer media inputs raises questions about multimodal preprocessing pipelines, context window management, and privacy handling for user-uploaded content.
What happened
Google began rolling out an update to Gemini on Android via Google app version 17.32, according to NokiaPowerUser. The update adds several new entry types to Gemini's Add menu: video clips, music files, projects, guided learning sessions, and other supported content types. NokiaPowerUser also reports that many Gemini capabilities can now be accessed from a floating interface without opening the full Gemini app, letting users ask questions, upload content, and continue conversations while remaining in other apps.
Context
The June 2026 update builds on earlier changes to the Gemini Android experience. In March 2026, 9to5Google reported that Google rolled out a redesigned Gemini overlay (stable v17.8) giving users access to a full Tools menu - including Deep Research, Canvas, Create image, Create video, Create music, and Guided Learning - from anywhere on Android. How-To Geek's coverage of the June 2026 Pixel Drop (released with Android 17) notes Google brought Gemini Omni video generation and Lyria 3 music generation to Pixel devices that month, consistent with video and music appearing in Gemini's Add menu via v17.32.
For practitioners
Adding video and music as assistant inputs typically requires substantial changes to the ingestion pipeline: audio-to-text transcription, video frame selection, cross-modal embedding, and additional safety passes. The floating, persistent interface implies ongoing session state, which raises engineering questions around context window management and what data is retained or uploaded between turns. Developers building on Gemini APIs should watch for new SDK surface area tied to these media types.
What to watch
Track developer-facing API and SDK updates that expose the new input types, changes to app permissions for camera and audio access, and any guidance Google publishes on data retention for uploaded media. Also watch for battery and performance impacts from persistent session handling on mid-range Android devices.
Scoring Rationale
A notable Android product update expanding Gemini's multimodal inputs and OS-level integration, relevant to platform engineers and ML practitioners tracking assistant UX evolution. Not a frontier model release; impact is mid-tier consistent with a product update that deepens an existing assistant's capabilities.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
