Gemini Tasks Opens to Free Users in Six Countries, Including Japan; Calling and Shopping Features Launch in the U.S.
On November 17, 2026, Google began offering Gemini Tasks to free users in Japan, the United States, South Korea, Canada, Australia, and India, enabling long-running projects and recurring tasks to be executed in the background. The free version allows up to 10 active schedules at a time and only supports specifying a time window for execution; Google AI Pro and Ultra support exact times, conditional triggers, and up to 50 schedules. U.S. users can also have Gemini use Google Duplex to call businesses or, with user authorization, purchase products through Google Pay.
Why read this
SelectedAI score 90
The differences between the free and paid versions in scheduling precision, trigger conditions, and task limits provide concrete criteria for evaluating Gemini automation workflows.
Topics
Body · AI translation
On November 17, 2026, Google began gradually making “Tasks” in the Gemini app available to people without a paid subscription in six countries, including Japan. For example, if users ask it to plan a trip, Gemini will put together an itinerary in the background and notify them when it is ready. Users can also entrust it with tasks such as checking their email and calendar every morning and summarizing the day’s priorities. In addition, features that let Gemini call stores on users’ behalf or purchase products with Google Pay will be rolled out gradually, beginning in the United States.
Put Gemini to work for you - Google
In an announcement in October, Google introduced a feature that allowed Google AI Pro and Ultra subscribers to request tasks from the chat screen. It has now expanded availability to free-plan users in Japan, the United States, South Korea, Canada, Australia and India. According to Google, hundreds of millions of people across these six countries will be eligible for free access. The free version of Tasks uses the “Gemini 4 Hydrogen” AI model.
Put Gemini to work for you with Tasks.
While Chat is for quick back-and-forth answers, Tasks are for longer projects and recurring schedules.
Beginning today, Tasks are now free of charge for hundreds of millions of eligible users, starting in the U.S., JP, KR, CA, AU & IN. 🧵— Google Gemini (@GeminiApp) November 17, 2026
How to request tasks and the requirements for using them
The feature is available to personal account users aged 18 or older and cannot be used with work or school accounts. To use it, users must sign in to the Gemini mobile app with a personal Google account and turn on “History” in the Gemini app. Once these requirements are met, users can request tasks from the web and Mac apps as well as the mobile app.
In the web version, users select “Task” from the “Chat” and “Task” options displayed below the input field. Google’s help documentation advises users to specify what they want completed and by when. When requesting research, for example, users can say, “Compile the results in a spreadsheet together with the sources.” They can add further instructions even after work has begun. On some sites, Gemini may hand control back to the user partway through and ask them to log in or perform the final button click themselves.
Integration with apps also enables sending emails and adding calendar events
In response to a request, Gemini uses information from apps that the user has connected to Gemini and from websites where the user is signed in. When checking a schedule, for example, it consults the connected calendar. Even with a personal Google account, users can open “Connectors” under “Customization” to see which apps are connected or disconnect them.
In a post on X, Google’s Josh Woodward explained that this update enables Gemini to invoke Google Workspace tools more reliably. Among the actions users can ask Gemini to perform, he listed drafting and sending emails, creating and editing Google Docs and spreadsheets, and adding multiple events to Google Calendar in a single operation.
How Gemini handles information when executing tasks
When the author began using Tasks in the web version, a notice appeared stating that Gemini might share information such as names and contact details with third parties in order to carry out the requested work. Users can stop a task in progress or switch to manual operation at any time[1].
According to Woodward, Google will also gradually roll out integration between Gemini and Google Password Manager, beginning in the United States. If users authorize this integration, Gemini will reportedly be able to sign in to websites without sharing passwords outside Google.
According to Google, Gemini processes tasks in a virtual machine dedicated to each user’s account. Each virtual machine is isolated at the hardware level from other users’ execution environments.
The free tier lets users specify a time window for running tasks, while Pro and Ultra support exact times
On the free tier and Google AI Plus, if users specify a time window, such as “Summarize tomorrow’s schedule every evening,” the task will begin at approximately the same time within that window. The shortest available recurrence interval is “daily,” with the task running once during the selected time window. Up to 10 schedules can be active at the same time.
Using the free web version, the author asked Gemini to compile a list every morning at 8:00 a.m. of articles published on gihyo.jp the previous day (at that point, the author did not know that the free tier did not support specifying an exact time). The progress panel displayed “Daily (around 8:40 a.m.)” as the estimated execution time.
Figure 1 When asked to run the task “every morning at 8:00 a.m.,” Gemini replied that it would run “daily, during the daytime window.”
Figure 2 After creating a task, pressing “Done” in the upper-right corner displays its progress. Pressing “Summarize gihyo.jp articles” in the “Schedule” section opens the task’s editing screen shown in the figure. The “Plan” lists the steps for finding and summarizing the articles. Under “Repeat,” users can select “Daily,” “Weekly,” “Monthly,” “Yearly,” or “Once,” while under “Time,” they can select “Next day,” “Daytime,” or “Evening.” “Daily” and “Daytime” are selected on the screen.
Figure 3 When asked to test the task immediately, Gemini replied that it had found four articles published the previous day, November 18, and organized their titles and summaries in a table.
For Pro and Ultra, receiving an email can also serve as a trigger
According to Google’s help page on scheduling, Google AI Pro and Ultra allow users to specify an exact execution time, such as “every morning at 7:00 a.m.” Users can also configure the same task to repeat “every hour.” The limit on active schedules increases to 50.
In addition to a time, a specific condition can be used to trigger a task. The same help page gives the example of having Gemini handle a request and draft a reply when an email containing a work request from the user’s manager arrives. Users can also ask Gemini to notify them when a local event is announced and provide its details. However, according to Google, Gemini is not yet optimized for tracking frequently changing data or sending notifications immediately after tickets go on sale.
In the United States, it can also make calls to businesses and shop on users’ behalf
Gemini uses Google’s automated calling technology, Google Duplex, to call businesses[2]. Woodward asked Gemini to call two nearby dealerships to schedule maintenance for his personal vehicle. According to the account in his announcement blog post, Gemini used the vehicle model and his home location to find dealerships on Google Maps, then called two of them simultaneously. It reportedly booked maintenance at one dealership for early Friday morning, when Woodward was available, and added the appointment to his calendar.
Gemini can now call local shops and sit on hold for you to book appointments — and then add the date and time directly to your Google Calendar. 📞 pic.twitter.com/9rTI9dRssn— Google (@Google) November 17, 2026
When requesting such a call, users can review in advance which business Gemini will call and what it will ask, then approve the call. Once connected, Gemini tells the other party that it is an AI. After the call ends, users can read a summary and the full transcript in the chat. Businesses can opt out of receiving calls from Gemini.
In the United States, users can also ask Gemini in English to shop for them. For example, they can specify the desired product and a spending limit, and instruct it to wait for a price drop or restock before making the purchase. Payment can be made using a card registered in the user’s Google Wallet or another supported method.
According to the November 17 release notes, under the default settings, when Gemini finds a desired product within the budget, it sends the order details to the user. Once the user approves them on their device, Gemini completes the payment through Google Pay. If the user turns off confirmation before purchase when requesting the shopping task, Gemini will buy a product matching the specified conditions without asking for approval again once it finds one.
In the United States, requests can also be submitted by message or email
In the United States, users can request tasks by sending Gemini an English-language text message (RCS) or email. To submit a request through Google Messages or iMessage, users must enable text-message requests in Gemini’s settings and complete verification using a U.S. mobile phone number. This feature is limited to lines operated by supported carriers. Even after Gemini starts working, users can add instructions by replying to Gemini in their messaging app.
To submit a request by email, users must sign in to both Gemini and Gmail with a personal U.S. Google Account and enable email requests in Gemini’s settings. For example, when forwarding a school event notice to gemini@google.com, users can ask in the email for the event date to be added to Google Calendar. After creating the task, Gemini replies by email with a link that opens it[3].
Text Gemini your tasks on Android and iOS, or forward emails directly to Gemini to create a Task or a Calendar entry with the information provided. 📅 pic.twitter.com/1GJTitBLwS— Google (@Google) November 17, 2026
AI Trends