Google said on November 17, 2026, that tasks in Gemini are rolling out free of charge in the United States, Japan, South Korea, Canada, Australia, and India, using the Gemini 4 Hydrogen model. In a blog post by Josh Woodward, Vice President, Labs, Gemini app, and AI Studio, Google attributed that reach to “hundreds of millions” of people, and said it is also rolling out “a set of advanced features, starting in the U.S.” The Verge’s Emma Roth reported the same day that “Google is rebranding its Spark AI agent as “Tasks” and making it free to use.” Google’s post does not mention Spark.

#What changed?

The Verge reported that Google first introduced its Spark AI agent in May, and that it was only available to Google AI Pro and Ultra subscribers. Google describes the feature as tasks in Gemini, and says people switch the toggle from Chat to Task in the Gemini app. Google says it is “updating Gemini to handle complex tasks like scheduling appointments, managing bills, and shopping for groceries.” The Verge reported that the feature can create meal plans, manage a calendar, and search email for recurring expenses. Google also said Gemini connects with more than 60 popular apps, and that developers can apply to join its Connected Apps Platform.

Google said the advanced features start in the U.S., and that people there can also forward emails to gemini@google.com or text Gemini. The Verge reported a wider U.S.-only set: the ability for the agent to call businesses to compare product pricing and check inventory, purchase items with Google Pay, and sign into accounts that support Google Password Manager, Google’s own calling example, described by Woodward, is booking a service appointment at nearby dealerships. Google says Gemini disclosed to the business that it is an AI, that businesses can opt out, and that it will not give out payment or private account information unless the user allows it. A footnote on the post says calling through Gemini is not available in Indiana, Louisiana, Minnesota, Montana, and Nebraska. For texting, Google says: “You must have a US phone number with an eligible mobile carrier.”

Google says Gemini uses the Agent Payments Protocol (AP2) to make purchases. It said it is working with American Express, Mastercard, and Visa to keep purchases secure with digital tokens on supported cards, and it did not give a date for American Express support. Google also says people can set a hard spending cap upfront, and that Gemini is designed to stay within that budget and to ask for confirmation or pre-approval before clicking “Buy”. A Gemini Apps help page says Gemini is designed to always ask for confirmation before making purchases and completing financial transactions, even if permission preferences are changed. Google said receipts and tracking information go to Google Wallet, which it said was expanding to iOS as of the November 17, 2026, post.

Google says assigned tasks run in Google Cloud’s Agent Compute Engine, in a “hardware-isolated virtual machine that’s dedicated to your account and never shared with other users.” The same post says Google has not eliminated all mistakes. The help page states that “Gemini’s ability to handle tasks is experimental”. It tells people not to enter sign-in information, payment details, or other sensitive information in the task thread.

#Who can use it and when?

A Gemini Apps help page says tasks are “Available without a Google AI subscription in Australia, Canada, India, Japan, South Korea, and the US.” People must be 18 or over, sign in with a personal Google Account, and have History on. The feature is not available with a work or school Google Account, and Google did not give a date for those accounts. It is “Only available in the Gemini mobile app, the Gemini app on Mac, and the Gemini web app at gemini.google.com.”

Outside those six countries, the help page says tasks are available with a Google AI Pro or Ultra subscription everywhere else Gemini Apps are supported, except in the European Economic Area, Nigeria, Switzerland, and the United Kingdom.

Plan or market

What Google’s help pages and November 17, 2026, post say

Free access in six countries

No Google AI subscription required. Standard limits. Up to 15 tasks at once and 10 active schedules, at most once a day in an overnight, daytime, or evening window.

Google AI Plus

Limits 2x higher than standard. Same 10-schedule cap and once-a-day windows as the free tier.

Google AI Pro or Ultra

Limits 4x higher than standard on Pro, and 5x or 20x higher than Pro limits on Ultra, depending on the subscription. Up to 50 active schedules, and three schedule types: time-based, Gmail monitors, and topic monitors. In other supported markets, a Pro or Ultra plan is required, with the exceptions listed above.

Advanced features

Google named email and texting as starting in the U.S. The Verge also listed calling, Google Pay, and Google Password Manager sign-in. Calling is not available in Indiana, Louisiana, Minnesota, Montana, and Nebraska. Texting requires a U.S. phone number with an eligible mobile carrier.

A limits help page says the usage limit refreshes every 5 hours until the weekly limit, and that if capacity changes, users without a Google AI plan may be limited first. Google does not publish an exact free-tier task quota beyond “Standard limits”. Google says people can open Gemini and flip the home-screen toggle from Chat to Task. In the U.S., the post also lists a texting setup under Settings, then Channels, then “Text with Gemini”, with phone-number verification.

#How does it compare?

Google says the free update uses its Gemini 4 Hydrogen model. A Gemini Apps help page says tasks can use Connected Apps and skills, which ties the feature to Google’s Connected Apps Platform preview and to the transition from Gems to skills. The U.S. texting and Google Pay path sits alongside other agent shopping tools, including Alexa for Shopping’s text-to-shop feature.

Our analysis: The main change is availability, not a brand-new agent. In six countries, Google is offering tasks in Gemini without a Google AI subscription, though usage still counts against compute-based limits that rise with higher subscriptions. The deeper agent features The Verge listed, including calling and Google Pay, are a U.S. rollout, and calling is not available in five states. Google’s post never uses the Spark name; that history comes from The Verge. On purchases, Google describes confirmation steps and a spending cap, and does not claim that checkout is error-free.

#What is still unclear?

Google has not said when the U.S. advanced features will reach the other five countries, or when tasks will be available with work or school accounts, or in the European Economic Area, Nigeria, Switzerland, and the United Kingdom. It did not give a date for American Express support, or for integrating the Universal Commerce Protocol. The help pages do not publish an exact daily task quota, or prices for Google AI Plus, Pro, or Ultra. Google’s November 17, 2026, post also does not describe who could use Spark before this rollout.