Google Expands Gemini Spark's 24/7 Background Agent to More Countries
Google's always-on AI agent Gemini Spark expands internationally with four new productivity upgrades including Google Docs editing and parallel source retrieval
- International expansion: Gemini Spark is now rolling out to Google AI Ultra subscribers in more countries and languages beyond the initial US launch.
- Four new upgrades: Google Docs editing, Sheets/Slides comment reading, speed improvements, and parallel source retrieval are all starting to roll out today.
- Still excluded: Australia, Canada, the EU, Hong Kong, India, Japan, Nigeria, South Korea, Switzerland, and the UK remain blocked -- EU/UK pending AI Act review.
- Pricing gate: Spark requires Google AI Ultra at $100/month minimum; no standalone purchase option exists.
- Competitive context: At $100/month, Google AI Ultra is the only plan in that tier to include a persistent 24/7 cloud agent, vs. OpenAI and Anthropic's per-task models.
- Architecture: Spark runs on isolated ephemeral VMs on Google Cloud via the Antigravity harness, continuing to execute tasks even when your device is off.
Gemini Spark, Google's 24/7 background AI agent, is now rolling out to Google AI Ultra subscribers in more countries and languages -- and it's shipping four meaningful capability upgrades alongside the expansion. This is not a soft announcement. It's the clearest signal yet that Google is treating Spark as the centerpiece of its AI subscription strategy.
From chatbot to agent: what Spark actually is
Spark represents a big shift for Gemini, transforming it from an assistant that can answer your questions into an active partner that does real work on your behalf and under your direction. The key architectural difference from a regular chatbot: because it is a cloud-based agent, Spark keeps working in the background even when you close your laptop or lock your phone.
Gemini Spark runs 24/7 on dedicated virtual machines in Google Cloud, powered by Gemini 3.5 and the Antigravity harness -- an agentic framework that allows it to perform long-horizon tasks in the background. Think of Antigravity as the scaffolding that lets Spark plan, execute, and recover from multi-step workflows without you babysitting it.
The agent operates on a three-layer model: Tasks (what you want done), Schedules (when or how often), and Skills (reusable personal work patterns you teach it). Spark proactively sends critical updates and requires explicit approval for high-risk actions like sending emails. The more you use it, the better it learns your unique preferences to become a more accurate, helpful extension of your work style.