Google has made Gemini Spark, its agentic AI assistant, available on macOS, significantly expanding its desktop AI capabilities. The macOS launch enables Spark to work directly with files on the computer and handle remote tasks across apps.
The update includes several notable enhancements:
Expanded App Integrations: Google has added support for Google Tasks and Google Keep — two integrations notably absent at Spark's initial launch. Third-party app support now includes Canva, Dropbox, Instacart, OpenTable, and Zillow Rentals, enabling Spark to perform diverse tasks like designing flyers, ordering groceries, reserving tables, and booking apartment tours.
Custom MCP Support: Users can now connect their favorite apps directly into Spark via custom Model Context Protocol (MCP), allowing for highly personalized AI assistant configurations.
Real-Time Topic Tracking: Spark can now track topics and react to events in real time, improving performance for tasks involving sports scores, stock movements, or breaking news.
Cross-Device Task Assignment: While not available at launch, Google plans to allow users to assign multi-step tasks across devices — for example, having Spark on a phone instruct the desktop agent to pull information from a file on the Mac.
Currently, Gemini Spark for macOS (beta) is available only to Google AI Ultra subscribers in the U.S. This expansion puts Google's agentic assistant in direct competition with Claude Desktop, Microsoft's Copilot, and OpenClaw on the desktop.