Skip to content
Agentic Levels
Start hereLevelsTracksPromptsAbout
Get my levelGet my level

  • Start hereYour first useful task with AI.
  • Level quiz8 questions. No AI needed.
  • Fluency ScoreYour Level and a score, from your own AI.

  • LevelsThe 11 levels, in order.
  • LessonsShort lessons, filter by level.
  • TracksGuided paths by tool, role or goal.
  • Set up your AI tool
  • Starter packFree files that help AI remember.

  • PromptsReady prompts to copy and adapt.
  • ChatGPT vs Claude vs GeminiPick the right one for the job.
  • How-tos
  • ReferenceCommands for Claude Code, Codex, Antigravity.
  • ResourcesHand-picked reading and tools.

  • Your progress
  • Your next moveAn interview that picks your next skill.
  • Your next lessonsThe next lessons at your level.
  • Updates

  • Pricing
  • About
  • FAQ & Feedback
  • Work with us
  • How we handle your data
Sign inPreferences
AboutPricingHow-tosPromptsStarter packMy stackHow we handle your data

© 2026 Fuentes Studio

PrivacyTermsRefundsAccessibilityAI content policyCredits
yourCouncil
Ready to help
✦

What do you want to understand?

Ask anything about what you're learning.

←Codex Fundamentals
  1. Tracks
  2. ›Codex Fundamentals
  3. ›How-to guide

How to use Gemini in Chrome

Google's Gemini, built into Chrome. Ask about the page you're on, compare up to 10 tabs, and on some paid plans let it click through a chore for you.

Levels 3 to 5Official page

You already live in Chrome. Copying a page into a chat window to ask about it gets old fast. Gemini in Chrome sits inside the browser, so it can read the page and your other tabs with you. With auto browse, it can also click through a multi-step chore, asking before anything important.

Before you start

You'll get more from this once you've done Approve actions by risk. It shows how to decide when an AI action can run, when it needs your approval, and when to block it, which is the call you make on every auto browse task.

  1. 1Check you can use it
  2. 2Turn it on
  3. 3Decide what it shares
  4. 4Ask about the page you're on
  5. 5Compare across your open tabs
  6. 6Save a request you'll reuse as a skill
  7. 7Hand a chore to auto browse, with a stop point
  8. 8Watch it, and take over when you need to

Steps

01

Check you can use it

Gemini in Chrome is built in, so there's no extension to install, but Google is still rolling it out by country, age, and language. It doesn't work in Incognito windows (Chrome's private browsing) or in a Chrome profile (one set of sign-ins) that isn't signed in. On a work or school account, your admin has to turn it on.

02

Turn it on

Sign in to Chrome and update it to the latest version. Click Ask Gemini at the top of the browser window and follow the steps to opt in. From then on, the same button opens Gemini next to the page you're reading.

03

Decide what it shares

Gemini shares the tab you're on by default. To change that, open the three-dot menu, then Settings, AI innovations, and Gemini in Chrome. That's where you switch off sharing the current tab by default, precise location, the microphone, and letting Gemini browse for you.

04

Ask about the page you're on

Start with a question about one page, so you can see what it picks up and what it misses.

Ask Gemini this
Explain this page for someone with [YOUR_CONTEXT]. Give me the three things that matter most, and tell me anything on the page I should double-check before I act on it.
[YOUR_CONTEXT]Who you are, in a few words e.g. a first-time landlord with no legal background
05

Compare across your open tabs

You can share up to 10 open tabs with Gemini. Type @ in the Gemini panel or use Add tabs, then ask it to compare them.

Ask Gemini this
Compare the [WHAT_THEY_ARE] in these tabs on [WHAT_MATTERS]. Put it in a table, mark anything a page doesn't say as unknown, and tell me which one fits [YOUR_SITUATION] and why.
[WHAT_THEY_ARE]What the tabs are e.g. four laptop listings
[WHAT_MATTERS]What to compare e.g. price, weight, battery life, and return policy
[YOUR_SITUATION]Your situation e.g. someone who travels every week and edits photos
06

Save a request you'll reuse as a skill

A skill is a saved request you can rerun on any page. In the Gemini panel, type /, click Add Skill, give it a name, paste in the request below, and click Save. Next time, type / and pick it.

Paste this into your skill
For this [PAGE_TYPE], pull out [WHAT_TO_PULL]. Keep it under [LENGTH], and flag anything that looks unusual or missing.
[PAGE_TYPE]The kind of page e.g. job posting
[WHAT_TO_PULL]What to pull out e.g. salary range, location, must-have skills, and how to apply
[LENGTH]How long e.g. six bullet points
07

Hand a chore to auto browse, with a stop point

Auto browse is rolling out gradually on paid Google AI plans, for adults in the US with a personal Google Account and Chrome in English, on a computer. It lets Gemini click through a multi-step task for you, like comparing products or checking times for a booking. Describe the task, read the plan Gemini shows you, then click Start Task.

Ask Gemini this
Find [WHAT_YOU_NEED] on [SITES]. Get it ready, then stop before [FINAL_STEP] and show me what you chose and why. Don't sign in anywhere new or send anything.
[WHAT_YOU_NEED]What you need e.g. a table for four this Friday between 7 and 8 pm, near downtown
[SITES]Where to look e.g. the booking sites I'm already signed in to
[FINAL_STEP]The step you'll do yourself e.g. confirming the reservation
08

Watch it, and take over when you need to

Gemini asks before sending messages, changing your data, submitting forms, scheduling events, or opening sites with sensitive financial or health data. It hands the task back to you to finish a payment, accept terms, or create an account. Click Stop in the side panel, or Take over task, whenever something looks off.

How do you know it worked?

  • Gemini opens next to a page when you click Ask Gemini
  • You've checked what it shares under Settings, AI innovations, Gemini in Chrome
  • You've compared at least two tabs and got a table back
  • You've saved one request as a skill you can run with /

If something looks off

You're responsible for what auto browse does
Google's help page says you're responsible for Gemini's actions during a task, including mistakes like an unwanted purchase. Put the stop point in your request and watch the run. Google calls watching your tasks the most important protection.
Forgetting your tab is shared by default
Gemini sees the tab you're on unless you turn that off. Before you ask about something else, check which tabs are shared and remove any you don't want it to read.
Hidden instructions on a page
A web page can carry text meant to steer Gemini off your task (prompt injection). Google says its safeguards don't remove every risk, so keep auto browse to sites you trust.
It may not be available to you yet
Availability depends on your country, age, and device language, and Google says it isn't available to everyone yet. Update Chrome, and turn off any VPN (a tool that can make Chrome look like it's in another country). Then check Google's availability page, or on a work account, ask your admin.
Closing the laptop mid-task
Google says an auto browse task in your own browser needs your computer and Chrome awake. Keep Chrome open until Gemini reports back.

See also

  • Claude in Chrome: Anthropic's version
  • The ChatGPT browser extension: OpenAI's version

What you'll do, 8 steps

  1. 1Check you can use it
  2. 2Turn it on
  3. 3Decide what it shares
  4. 4Ask about the page you're on
  5. 5Compare across your open tabs
  6. 6Save a request you'll reuse as a skill
  7. 7Hand a chore to auto browse, with a stop point
  8. 8Watch it, and take over when you need to

Was this helpful?