Google Connects Gemini Live to Spark for Voice Tasks

Google has upgraded its Gemini Live voice assistant by connecting it directly to Spark, the company's persistent background agent framework. This integration allows users to initiate complex, multi-step workflows across Google Docs, Sheets, Drive, and the web using spoken commands. Instead of acting as a simple conversational chatbot, Gemini Live now serves as an orchestrator that can hand off long-running tasks to Spark, which executes them silently in the background.

Under the hood, Spark runs on Google's Gemini 3.7 Flash model and can manage up to 15 concurrent tasks at a time. The update also introduces Daily Brief, which synthesizes updates from Gmail and Calendar into a single spoken digest, and Personal Intelligence, a long-term memory layer that links context across past chats, Photos, Search, and YouTube. For safety, Spark is designed to request explicit user approval before executing sensitive actions like sending emails or making purchases.

While Gemini Live remains free globally, accessing these advanced agentic features requires paid tiers. Spark is available to Google AI Pro and Ultra subscribers in more than 160 countries, excluding the European Economic Area, the United Kingdom, Switzerland, and Nigeria. Meanwhile, the Daily Brief feature requires at least a Google AI Plus subscription. Personal Intelligence and hands-free Gmail controls are managed through the connected apps settings in the Gemini application.

For developers and productivity-focused users, this shift represents a transition from basic voice commands to true autonomous delegation. By combining a low-latency voice interface with a persistent cloud agent, Google is attempting to make voice interactions practical for complex, multi-day logistics rather than just quick queries. For those heavily reliant on the Google Workspace ecosystem, this update turns Gemini Live into a functional digital assistant capable of handling tedious administrative tasks.