Gemini Tests Mobile AI Agent Creation for All Users
Google appears to be testing a significant expansion for its Gemini AI. Evidence from the latest Android app version suggests the company is bringing agent creation capabilities directly to phones and tablets. If it launches, this would be a major shift. Custom AI agents are currently restricted to web interfaces and only available to Google Workspace and Enterprise users.

The Mobile Agent Creator Emerges

Code discovered in version 17.45.14 of the Google app for Android reveals a New agent option sitting directly in Gemini’s sidebar, right beneath recent chats. This discovery comes from an APK teardown — the type of dig that flags work-in-progress features that may or may not ever ship publicly.

Currently, agent creation is limited to Google Workspace and Enterprise users, accessible only through a web interface. Mobile access would change that equation entirely.

Five Ready-Made Templates to Start With

Instead of free-form customization like the web version offers, the mobile interface appears to start with five pre-built templates. These templates aim to get users creating agents quickly for common personal and professional tasks:

  • Family coordinator: Helps plan household chores and manage family events
  • Tutor: Assists with study topics, homework, and learning progress
  • Group organizer: Facilitates management of outings and gatherings
  • Marketing agent: Supports social media posts, newsletters, and campaign efforts
  • Finance agent: Tracks spending and manages bill payment reminders

The template-first approach suggests Google is optimizing for speed over flexibility. Rather than asking users to describe what they want an agent to do, Google is offering proven starting points. This is a smart design choice for mobile, where typing long descriptions on a small screen isn’t ideal.

What This Signals About Google’s Plans

The personal nature of these templates suggests something important: Google is preparing to extend agent creation well beyond Workspace and Enterprise tiers. Wider availability for all Gemini users, including non-paying ones, seems not just plausible but likely.

This aligns with Google’s broader mobile strategy. The company has also been testing other Gemini upgrades on Android, including the ability to customize Daily Briefs and manage skills for Spark, its always-online agent. Mobile AI customization is clearly becoming a priority.

The Open Questions

Two things remain unclear. First, will the mobile interface eventually offer custom descriptions like the web version, or will it stay limited to templates? Second, when does this actually launch? APK teardowns reveal future features, but they don’t guarantee a public release. Google could test this feature and quietly abandon it.

What we know for certain is that Google is moving in the direction of mobile-first AI customization. Whether that means everyday users will get full agent creation tools in the coming weeks or months remains to be seen.

Follow Hashlytics on Bluesky, LinkedIn, Telegram and X to Get Instant Updates