Google Gemini Spark Brings Persistent, Cross-Device Task Automation to Gemini

Google is unveiling Gemini Spark, a persistent AI agent designed to handle multi-step tasks across devices, as part of its June 2026 Gemini update. Unlike conventional assistants, Spark is intended to retain and execute workflows even when apps or devices are closed, moving beyond the momentary nature of a single chat response. The most significant distinction, according to Google, is that Spark can transform a conversationally defined task into prolonged background work.

While Gemini Live already offers voice-first interactions, Spark adds the capability to configure tasks for continuous operation. Google has demonstrated a workflow where a complex task initiated on a mobile device is completed on a macOS host computer while the user is away. The core capabilities of Spark include background execution, cross-device orchestration (enabling mobile-to-desktop handoffs), conversational task setup for creating tasks and reusable skills, and integration with Google Workspace.

However, the scope of supported actions, integrations, and task limits remains undefined in the provided information. Businesses are advised to evaluate the specific workflows Spark supports in their own environment rather than assuming it can handle all routine work. Gemini Live and Spark serve different roles, with Live being the voice-oriented interaction layer and Spark being the agentic execution layer that turns defined tasks into longer-running work.

Currently, Spark is available to Google AI Ultra subscribers aged 18 and over in the United States, with a limited rollout to Google AI Ultra subscribers aged 18 and over in the United States. Pricing information is not provided in the verified materials. For teams considering Spark, it is recommended to start with a bounded task with a clear outcome and review point, assessing whether Spark can complete steps reliably, whether tasks can be described conversationally, and whether the mobile-to-desktop workflow aligns with the team's working habits.

It is also crucial to establish which work should remain under direct human control. While persistent automation can streamline repetitive tasks, it should not replace human review of consequential work.