Google's decision to push Gemini Spark onto the desktop signals it now sees agentic assistants as an operating-system-level battleground, not just a chat feature, putting direct pressure on Claude Desktop and Microsoft Copilot as the default AI layer on knowledge workers' machines.
On June 30, 2026, Google announced Gemini Spark is coming to the Gemini macOS app in beta for Google AI Ultra subscribers age 18+ in the US, letting the assistant automate file-sorting, spreadsheet-building, and other desktop tasks with permissioned file access.
The update adds native integrations with Google Tasks, Google Keep, Canva, Dropbox, Instacart, OpenTable, and Zillow Rentals, plus support for custom Model Context Protocol connections so users can wire in their own tools.
Google also said Spark will soon accept multi-step tasks assigned remotely from a phone while a user's Mac executes them unattended, and added real-time topic tracking for sports, stocks, and news so the assistant can proactively notify users rather than wait to be asked.
Google's decision to push Gemini Spark onto the desktop signals it now sees agentic assistants as an operating-system-level battleground, not just a chat feature, putting direct pressure on Claude Desktop and Microsoft Copilot as the default AI layer on knowledge workers' machines. On June 30, 2026, Google announced Gemini Spark is coming to the Gemini macOS app in beta for Google AI Ultra subscribers age 18+ in the US, letting the assistant automate file-sorting, spreadsheet-building, and other desktop tasks with permissioned file access. The update adds native integrations with Google Tasks, Google Keep, Canva, Dropbox, Instacart, OpenTable, and Zillow Rentals, plus support for custom Model Context Protocol connections so users can wire in their own tools. Google also said Spark will soon accept multi-step tasks assigned remotely from a phone while a user's Mac executes them unattended, and added real-time topic tracking for sports, stocks, and news so the assistant can proactively notify users rather than wait to be asked.
