Gemini in any tab: Google's models in a browser sidebar

Pick a Gemini model and the sidebar answers with Gemini while it can see the page you are reading.

Long documents and mixed content

Copilotly's picker describes Gemini 3.1 Pro as built for long documents and multimodal work, so it suits pasting in a long report, asking about a page with images, or summarising a YouTube video through the video summary feature. Google lists Gemini 4 Argon with text, image, audio, video and PDF input.

The rest of the product

Whichever model you pick, you can rewrite selected text anywhere, translate a whole page bilingually, search the web with cited sources and generate images.

Switching models

If Gemini's answer looks thin, switch to Claude or GPT in the same conversation and compare.

Google builds Gemini into Chrome itself rather than as an extension. as of November 2026, Google's help page describes Gemini in Chrome as an assistant in the Chrome side panel that answers questions and summarises pages, using your current tab by default and up to 10 more open tabs on a computer.

What Google's version requires

• A Chromebook Plus, Mac or Windows computer with the latest Chrome.
• Being signed in to Chrome; it is not available in Incognito mode.
• A supported region and language. Google says it is rolling out gradually and may not be available to everyone yet.
• For work or school accounts over 18, an administrator must enable it.

The page covers desktop Chrome only. Google's separate Gemini apps and site are at gemini.google.com.

When Google's version is the better choice

If you live in Google's apps and want Gemini wired directly into them, Google's own integration reaches Google services in ways a third-party extension cannot.

Where Copilotly differs

• Gemini beside other models. GPT, Claude, Grok, DeepSeek and Mistral are in the same sidebar.
• Other browsers. The same extension works in Edge, Firefox, Brave and Opera, not only Google's browser.
• Not tied to a Chrome sign-in. You use a Copilotly account, and your browser's own per-extension setting decides whether any extension runs in private windows.
• Bring your own key. Use your own Google API key and pay Google directly.

Google's model page describes some of the Gemini models and not others (as of November 2026). Where Google says nothing, the description below is Copilotly's picker text.

Model	Tier	Free plan	Pro plan
Gemini 3.5 Flash Lite	fast	Included, within your token allowance	Included, within your token allowance
Gemini 3.8 Flash	fast	Included, within your token allowance	Included, within your token allowance
Gemini 3.1 Pro	premium	10 messages a day, then Gemini 3.8 Flash	Unlimited
Gemini 4 Argon	frontier	Your own API key only	30 a day, then Gemini 3.1 Pro

Gemini 4 Argon

Google calls it an efficient frontier reasoning model with faster responses and lower token spend per turn, aimed at complex coding, agentic, math and science tasks. It takes text, image, audio, video and PDF input with a 1M-token context window.

Gemini 3.1 Pro

Google's page does not describe it. Copilotly's picker lists it for long documents and multimodal work.

Gemini 3.8 Flash

Google calls it its most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents and complex enterprise workflows. Copilotly's picker describes it as fast, with a huge context window.

Gemini 3.5 Flash Lite

Google's fastest and most cost-effective 3.5 model, for high-throughput work. It is the lightest and quickest in the picker.

Pro is $6.99/week, $19.99/month or $149.99/year, with a 7-day free trial for first-time subscribers. Details on the pricing page.

On the free plan

Yes. The free plan includes both Gemini Flash models and 10 premium-model messages a day for Gemini 3.1 Pro. A free account is needed; a card is not. After 10 premium messages the chat continues on Gemini 3.8 Flash.

On your own Google API key

Every model on this page can also run on your own API key. Requests made on your own key are never metered by Copilotly, whatever the model's tier, and that includes frontier models on the free plan, which are otherwise Pro-only. You pay the provider directly at their API rates. The steps are on the bring your own key page.

Connect a Google API key and Gemini 4 Argon works on the free plan too, billed by Google at its API rates. Google's API billing is separate from any Google AI subscription you hold, so check how your Google account is set up first.

Big or mixed input

Reach for Gemini 3.1 Pro when the input is large or mixed: a long document, a page with images, a video. Use it with chat with PDF for a long report.

Reasoning-heavy work

Use Gemini 4 Argon on Pro for harder coding, math or science questions, which are the tasks Google names for it.

Everyday questions

The Flash models are quicker and available on every plan. Gemini 3.8 Flash is Google's pick for long-horizon agentic work; for this sidebar its main value is speed on ordinary questions. Gemini 3.5 Flash Lite suits very quick lookups.

Compare before you rely on it

If you are not sure which model suits a job, run the same prompt through Gemini and Claude and compare. Check anything important either one tells you; the ChatGPT vs Claude vs Gemini guide gives more examples.

Of the vendor pages we read, Google's is the one that lists audio, video and PDF input for Gemini 4 Argon alongside text and images. That is a real difference if your sources are recordings or documents.

Model	Context window	Image input	Source (Nov 2026)
GPT-6.1 Terra	1.05M	Yes	OpenAI docs
Claude Sonnet 5.5	1M	Yes	Copilotly catalogue
Gemini 4 Argon	1M	Yes, plus audio, video and PDF	Google AI Studio docs
DeepSeek V4 Pro	1M	No (Flash: yes)	DeepSeek docs
Grok 4.7	500K	Yes	xAI docs
Mistral Large 4	1M	Multimodal	Mistral docs

What the table does not show

Input types and context size say what you can send, not how good the answer is. Google's page frames Argon around efficiency (faster responses, lower token spend per turn) as well as capability, so test it against Gemini 3.1 Pro on your own work before assuming the newer one is better for you.

How to test fairly

Use the same prompt, the same page and the same instructions. Switch the model in the menu, not the wording. Compare the two answers on what you can verify, such as numbers you can find in the source.

Match the model to the input

For a long report, use Gemini 3.1 Pro with chat with PDF. For quick questions about the open page, a Flash model is faster and free of the premium limit.

Ask for structure

For a long document, ask for an outline first, then ask follow-ups about the sections you care about. That is easier to check than one long summary.

Verify numbers against the source

Ask which page or section a figure comes from, then look at it. Do the same for video: YouTube summaries are a way to decide what to watch, not a substitute for watching.

Mix in web search when the page is not enough

If a question goes beyond the page, turn on web search so the answer cites live sources.

Through Copilotly: Gemini in Firefox and Opera work, and Safari does not yet. Google's own Gemini in Chrome is a feature of Chrome on desktop, so it does not extend to other browsers.

Where	Status	How
Chrome	Supported	Chrome Web Store
Edge	Supported	Its own Edge Add-ons listing
Firefox	Supported	Its own Firefox Add-ons listing
Brave	Supported	Installs the Chrome Web Store build
Opera	Supported	Installs the Chrome Web Store build
Arc, Vivaldi	Supported	Install the Chrome Web Store build
Safari	Not yet	There is no Safari build in the App Store
iPhone, Android	Supported	The Copilotly iOS and Android apps, which work on the free plan

On an iPhone or Android phone, use the Copilotly mobile app.