Google is rolling out Gemini Tasks free of charge across the United States, Japan, Korea, Canada, Australia and India. In its November 17, 2026 announcement, the company says the update brings task execution powered by Gemini 4 Hydrogen to hundreds of millions of people.
The rollout has two parts. General task capabilities are expanding across all six countries. Advanced features, including calling businesses, purchasing through Google Pay and signing into accounts through Google Password Manager, are starting in the U.S. Getting free access to Tasks does not automatically mean getting every calling or payment feature. Google’s announcement and The Verge’s reporting make that distinction.
The practical change is that Gemini can move beyond explaining how to do something and carry out parts of the job. The important questions are which tools your account receives, which information you connect and what authority you give it to act.
What’s Free, and What Starts in the U.S.
The Verge reports that Google is rebranding its earlier Spark agent as Tasks and expanding access beyond its previous paid availability. General capabilities include creating meal plans, managing calendars and searching email for recurring expenses.
The more consequential additions sit in the U.S.-first portion of the rollout:
| Capability | Announced availability | Practical boundary |
|---|---|---|
| General Gemini Tasks access | Free rollout across the U.S., Japan, Korea, Canada, Australia and India | Access to Tasks is separate from access to every advanced tool |
| Business phone calls | Starting in the U.S. | Gemini can contact businesses for tasks such as checking availability and arranging appointments |
| Google Pay purchases | Starting in the U.S. | Supported payment methods and user authorization still matter |
| Google Password Manager sign-in | Starting in the U.S. | Intended for supported accounts, with permission |
| Assigning tasks by email or text | Included among the U.S. additions | These provide alternative entry points to the in-app task interface |
The regional breakdown comes from The Verge, which identifies the advanced features more explicitly than Google’s headline announcement.
Google describes this as a rollout, so country eligibility should not be read as a promise that every account immediately has every switch. Account and feature-specific availability may differ as access expands.
Gemini 4 Hydrogen supplies the underlying model, but a model rollout and a tool rollout are separate things. Access to Hydrogen elsewhere in Google’s products does not itself grant phone-call or checkout permissions. Zeniteq’s coverage of Gemini 4 Hydrogen across Search and the API explains that wider model release.
How to Switch From Chat to Tasks
Gemini exposes task execution through a toggle on its homepage. The Verge and TechCrunch both report the same basic setup:
- Open Gemini.
- Find the homepage toggle labeled Chat.
- Switch it to Tasks.
- Describe the job and any constraints.
A useful request specifies the outcome, timing and limits. For a service appointment, that might mean identifying the business, giving acceptable dates and saying whether Gemini may book a slot or should return options first. For shopping, it means specifying the item requirements and maximum spend.
Those are suggested instructions, not a guarantee that every workflow enforces each condition identically.
U.S. users also gain ways to assign work without opening the task interface. Google demonstrates forwarding school emails to gemini@google.com and sending shopping requests through its new RCS integration. In its grocery example, a follow-up photo identifies another item to add to an Instacart order. Google describes these entry points in its announcement.
The in-app toggle is the clearest place to start because it lets users check whether Tasks has reached their account before relying on email or messaging access.
What the Calls and Bill Examples Actually Show
Google’s dealership example combines several distinct operations: finding businesses, making calls, checking availability, choosing an appointment and updating a calendar.
In the company’s account, Gemini uses Personal Intelligence to identify the family’s vehicle and location, finds nearby dealerships through Google Maps, then calls two dealerships concurrently. It checks Google Calendar and books an early-Friday appointment. Google says Gemini identifies itself as AI at the start of the calls, businesses can opt out, and payment or private account information is not disclosed without permission. The example appears in Google’s announcement.
That workflow shows why connected context matters. The assistant needs the vehicle, location and scheduling constraints before making useful calls. It also shows that an authorized task can include an actual calendar change, without a separate confirmation for every intermediate step.
The bill example is narrower than the announcement’s broad bill-paying pitch. Google executive Josh Woodward describes granting access to a utility portal through Password Manager and asking Gemini to analyze earlier billing statements in email. He attributes a $389 electric bill partly to a heat pump operating during peak pricing.
That demonstrates account access and bill analysis. It does not, by itself, show a completed utility payment. Similarly, Google mentions help sorting medical bills without detailing a universal medical-payment workflow.
Google’s shopping examples show another distinction:
- A charger purchase uses Google Pay checkout assistance.
- An Instacart task adds groceries, then hands the user to Instacart to complete the order.
These are Google-reported experiences, not independent reliability measurements. TechCrunch’s coverage also attributes the dealership and electricity-bill examples to Google’s post.
I’d read each example as evidence of a particular integration, not as a promise that Gemini can finish any transaction involving a bill or shopping cart. Finding an amount due, diagnosing a charge, preparing checkout and submitting payment are different actions.
Connected Data and Approval Are Separate Controls
Google lists email, calendars, passwords, photos, files and payment cards as information that can support the task experience. Its examples use several specific connections:
- Email supplies statements, school messages and other task context.
- Google Calendar supplies availability and receives appointments.
- Personal Intelligence supplies relevant personal context.
- Google Password Manager enables permitted account access.
- Google Pay supports checkout.
- Photos can identify something the user wants added to an order.
These connections make an instruction more actionable, but they also expand what the assistant can reach. Google’s announcement says the user can provide instructions while a task is running and that Gemini is designed to pause before critical steps. It describes confirmation or pre-approval, while The Verge reports permission checks for sensitive actions such as purchases.
Frequently Asked Questions
4 questions
1Is Gemini Tasks free?
Google says it is rolling out free to hundreds of millions of people in the U.S., Japan, Korea, Canada, Australia and India. Advanced features have a separate U.S.-first rollout.