Using the Skills feature in Gemini Spark, you can train the agent by customising it to your liking. (Image Credit: Google)
Hyderabad: Google has officially announced the roll out of its AI-powered personal agent, Gemini Spark. Showcased at Google I/O 2026 event, the Gemini Spark is a 24/7 personal agent to automate tasks, which is powered by Gemini 3.5 AI model and the Antigravity platform. It runs on dedicated Google Cloud virtual machines, allowing Gemini Spark to perform long-running tasks in the background without requiring a device to remain active. The announcement marks a significant step in Google’s push towards ambient AI automation, addressing a growing demand from users who want their digital work handled without requiring them to be online or actively engaged with their device.
Who Can Access Gemini Spark?
Gemini Spark is currently rolling out in the United States (US) exclusively for Google AI Ultra subscribers. It is priced at $99.99 (approximately Rs 9498.32) per month. The agent is available across Android, iOS, and web platforms. On the web, it appears in a side panel alongside the Chat tab, while on mobile it sits between Search Chats and Daily Brief.
Gemini Spark is now available to all Google AI Ultra subscribers in the U.S.
It can handle the heavy lifting and connect the dots across your digital ecosystem to take action where it matters most. Whether you watch it work or let it run in the background, Gemini Spark is always… https://t.co/y0WzN813kB
— Google Gemini (@GeminiApp) May 29, 2026
Three Core Feature of Gemini Spark
Gemini Spark operates across three primary feature sets — Tasks, Skills, and Schedules — each designed to handle a different dimension of a user’s digital workload.
The Tasks feature integrates with Google Workspace tools including Gmail, Google Calendar, Docs, Sheets, and Slides. For instance, Spark could independently search for internship opportunities on a user’s behalf, acting on a single instruction without further input.
The Skills feature allows users to teach Spark personalised preferences for how specific tasks should be completed. A user could, for instance, configure a custom “Ghostwriter” skill that trains the agent to draft emails in their particular writing style consistently.
The Schedules feature enables time-based or conditional automation triggers. Users can instruct Spark to check their inbox every Monday morning at 9 AM, summarise key emails from the previous week, and generate a prioritised to-do list — all without manual prompting.
What is Coming Next
Google has confirmed that further capabilities are planned for Spark later this summer. Among the most notable upcoming additions is the ability for Spark to make purchases on a user’s behalf. Support for the Gemini desktop application is also expected to follow in due course.
Despite its autonomous nature, Google emphasises that Spark operates strictly under user instruction and will always seek permission before taking any significant action, this allows users to be in control of the agent’s activity.