Google is expanding access to Gemini Spark, its agentic AI assistant, to AI Pro subscribers in the US and AI Ultra subscribers globally as of July 24, 2026. The rollout moves the tool beyond its initial exclusive limit to the highest-priced subscription tier, though it remains unavailable in the UK, EEA, Switzerland, and Nigeria.
For months, Google’s most advanced automation tool was a luxury for a few. Gemini Spark, the 24/7 personal agent
unveiled at I/O 2026, was previously locked behind the expensive AI Ultra subscription. Now, Google is pushing the tool toward a broader user base, significantly lowering the barrier to entry for those in the US.
Subscription Tiers and Regional Access
The shift in availability represents a strategic move to move Spark from a limited preview to a more accessible productivity tool. In the US, the agent is now rolling out to those paying $19.99 per month for the AI Pro tier. This is a sharp drop from the original requirement of the AI Ultra plan, which starts from $100 per month (with some reports citing a $99.99 or $99 starting price).
While US AI Pro users are gaining access, AI Ultra subscribers are receiving the rollout globally, including local language support. However, this expansion is not universal. According to Google’s documentation, the tool is not available in the European Economic Area (EEA), the UK, Switzerland, or Nigeria.
Not everyone is getting a seat at the table. Spark remains unavailable to free users and those on the AI Plus subscription, which costs $7.99 per month.
| Subscription Tier | Monthly Cost | Spark Availability |
|---|---|---|
| AI Ultra | $99 – $200 | Global (Excluding EEA, UK, Switzerland, Nigeria) |
| AI Pro | $19.99 – $20 | United States (Other countries coming soon) |
| AI Plus | $7.99 | Unavailable |
| Free | $0 | Unavailable |
Automating Workflows via Skills and Schedules
Unlike a standard chatbot that simply answers prompts, Gemini Spark is an agentic AI. This means it can perform multi-step tasks in the background without constant user input. To manage this, Google introduced a framework of Skills (reusable instructions with context) and Schedules (automated triggers based on time or conditions).
For example, a user can set a high-level task to manage a business trip. If a flight is delayed, a specific schedule trigger can notify the user and propose an updated itinerary. To execute this, Spark can simultaneously use a “Travel Booking” skill and a “Gmail Writing” skill to rebook a room and send a confirmation email.
The system allows users to have up to 15 tasks running at a given time
, though it adheres to the same compute-based usage limits as the rest of the Gemini ecosystem.
Deep Integration with Google Workspace and macOS
Spark’s utility is tied to its ability to act across different apps. It is deeply integrated into Google Workspace, allowing it to search and summarize email threads, manage Calendar events, and create or edit documents in Docs, Sheets, and Slides.
The integration extends to other Google services as well. Spark can now connect to Google Tasks and Google Keep, allowing users to manage short lists and reminders without needing to create full documents.
A significant expansion occurred with the launch of Gemini Spark for macOS. For instance, Spark can sort files in a Downloads folder or turn local invoices into a budgeting worksheet in Google Sheets.
- Canva: Designing flyers.
- Dropbox: File management.
- OpenTable: Reserving tables.
- Instacart: Ordering groceries.
- Zillow Rentals: Booking apartment tours.
Real-Time Tracking and Custom Protocols
Beyond static tasks, Spark is evolving to handle dynamic data. Google has added intelligent topic tracking, enabling the agent to react to events in real time. This allows users to monitor stock movements, sports scores, and breaking news across blogs and social media.
To further personalize the experience, Google is rolling out support for the Model Context Protocol (MCP). This allows users to connect their own preferred apps directly into Spark, building a more tailored assistant.
This move places Google in direct competition with other desktop AI agents like Claude Desktop, Microsoft Copilot, and OpenClaw. While the ability to manage local files is a powerful productivity gain, it introduces new risks.
The rollout for AI Pro subscribers is currently limited to the US, but Google has promised that the feature will be available in other countries soon
.
Related reading
