Google’s Gemini Auto Browse can navigate websites, fill out forms, compare prices, and complete multi-step tasks without users handling every click themselves. However, access depends on subscription eligibility, device requirements, and Google’s rollout.
Google AI Pro and Ultra subscribers can use the feature in supported versions of Chrome, with access announced for the US and India. Standard Gemini in Chrome is available more broadly, but using Gemini to summarize a webpage does not automatically grant access to Auto Browse.
For eligible users, Google’s official instructions explain how to launch a task, approve an action plan, and take control when the browser needs assistance.
What you need to use Gemini Auto Browse
Auto Browse launched in preview for US desktop users in January 2026. Google subsequently announced Android support and expanded availability to India on Oct. 7.
According to Google’s India announcement, the feature is available on desktop and Android for eligible AI Pro and AI Ultra subscribers in that market.
Google’s general Help documentation lists these requirements for personal desktop accounts in the US:
- Age: At least 18 years old.
- Subscription: Google AI Pro or Google AI Ultra.
- Browser: The latest version of Chrome.
- Operating system: A compatible Windows, macOS, or Chromebook Plus device.
- Language: Device language set to English.
- Account: A signed-in personal Google Account.
- Security: Standard or Enhanced Safe Browsing enabled.
Auto Browse does not work in Incognito mode. Google’s documentation also excludes iPhone and iPad access.
Eligible work and school accounts may have separate administrator-managed access, so organizations should confirm their Workspace edition and Chrome policies before deploying the feature.
The rollout remains gradual. Meeting the listed requirements does not guarantee that the Auto Browse option has appeared in an account.
How to start an Auto Browse task in Chrome
Google documents the following procedure for eligible desktop users:
- Open the latest version of Chrome and sign in with an eligible account.
- Click Ask Gemini at the top of the browser.
- Enter a task you want Gemini to complete, such as comparing hotel prices for a business trip.
- Submit the request.
- Review Gemini’s proposed plan, including which websites it intends to visit.
- Click Start Task if the plan is correct.
- Monitor the task in Chrome and respond if Gemini requests confirmation or assistance.
Gemini displays an Auto Browse indicator on the tab where it is working.
You can click Take over task to complete a step yourself, then select Resume or Give back task to return control. You can also stop a task from the Gemini panel.
On supported Android devices, the process begins with the Ask Gemini control in Chrome. Google’s Android-specific documentation describes additional permission controls and a separate mobile interface.
Five tasks Gemini Auto Browse can handle
Google’s examples cover routine browsing, shopping, and administrative activities. These are supported use cases, although results depend on the websites involved and whether the necessary permissions are granted.
1. Compare travel prices and book accommodations
Auto Browse can research flights and hotels across different dates, compare options, and help identify less expensive travel arrangements.
For example, a user could request hotel options near a conference venue within a particular budget.
Google also describes using Auto Browse to book event parking through services such as SpotHero.
2. Fill out lengthy online forms
Auto Browse can enter information into supported web forms, reducing repetitive typing.
Google has demonstrated form completion using information extracted from documents, including PDFs.
Users should review the information before submission. Google identifies submitting web forms as an action that may require explicit confirmation.
3. Manage subscriptions and recurring orders
Google has demonstrated Auto Browse changing a recurring pet-food order through Chewy.
An eligible user could ask the browser to locate an existing subscription and update an order based on new requirements.
Changes affecting an account or payment should be reviewed before completion.
4. Shop for products and groceries
Auto Browse can compare products, search for discounts, and add items to shopping carts.
Google’s demonstrations include identifying products from an image and finding similar items within a specified budget.
Its October India announcement also identifies weekly grocery orders as a supported example.
Adding items to a cart does not necessarily authorize payment. Google says the system is designed to request confirmation before certain sensitive actions, including purchases.
5. Handle appointments and routine administration
Google lists scheduling appointments, collecting tax documents, filing expense reports, and checking bills among activities tested with Auto Browse.
Users can also ask it to find restaurant reservations or gather information across multiple websites.
However, these examples do not establish compatibility with every booking system, payment provider, or business application.
Auto Browse limits, permissions, and privacy controls
Google publishes daily request limits for the feature:
| Subscription | Auto Browse requests per day |
|---|---|
| Google AI Pro | Up to 20 |
| Google AI Ultra | Up to 200 |
The service also limits the number of tasks that can run simultaneously.
Auto Browse can access websites where a user is already signed in. With permission, it can also use Google Password Manager to complete sign-in steps.
Google says passwords are not directly shared with Gemini. Users can review permitted automatic sign-ins through Chrome’s settings.
To manage these permissions:
- Open Chrome Settings.
- Select AI innovations, then Gemini in Chrome.
- Under Permissions, find Sites Gemini can sign you in to.
- Remove authorization for individual sites when necessary.
The same settings area includes Let Gemini browse for you, which can be switched off.
Organizations evaluating credential access should distinguish these automatic sign-in permissions from broader password and passkey management policies.
What actions require confirmation?
Google says Auto Browse is designed to ask for confirmation before sensitive operations, including purchases, sending communications, changing data, and submitting forms.
It can also require users to take over for actions such as accepting terms of service, creating an account, or completing a financial transaction.
These protections do not guarantee that every action will be performed correctly.
Google specifically warns about mistakes, unexpected purchases, and prompt injection attacks in which malicious webpage content attempts to influence an AI agent.
Organizations using managed Chrome environments can configure Auto Browse access through administrator policies, including controls over whether the feature is enabled and which websites it can use.
This complements broader enterprise efforts to control AI automation across managed devices and applications.
What to do if Auto Browse is missing or not working
If the feature does not appear, first confirm that Chrome is updated and that the account has an eligible subscription.
Check the device language, geographic availability, and Safe Browsing settings. Auto Browse requires Standard or Enhanced protection rather than disabling Safe Browsing.
Next, open Settings > AI innovations > Gemini in Chrome and check whether Let Gemini browse for you is available and enabled.
For work or school accounts, contact the administrator to confirm whether the organization’s edition and policies permit Auto Browse.
If all requirements are satisfied but the feature remains unavailable, the gradual rollout may not have reached the account.
Google’s Chrome feature documentation also explains how to enable and use the regular Gemini assistant, which should not be confused with Auto Browse eligibility.
Auto Browse is useful for repeatable tasks such as comparing travel options, completing forms, or managing orders. For sensitive activities, users should review Gemini’s plan, monitor the websites it visits, and verify the outcome before relying on the completed task.
Also read: Google’s five Chrome study features include Gemini-powered quizzes, media analysis, and research tools available separately from Auto Browse.