Is Google Gemini Free? Free Plan Explained (2026)

Yes, Google Gemini is free to use. The free plan costs $0 with no credit card, and it still covers image generation, voice conversations, and Deep Research. What changed on October 9, 2026 is the model: free accounts now run on Auto, which sends most prompts to Gemini Flash-Lite and may pass harder ones to Flash or Pro, replacing the Gemini 3.6 Flash and varying Gemini 3.1 Pro access the free tier had before. For a lot of people the free tier will still be all they ever touch, but it is a smaller one.

The catch is limits, not features. Free Gemini runs on usage caps that refresh every five hours inside a weekly limit, holds back video generation, and since October 9 leaves the choice of model to Gemini unless you pay, which is where the paid Google AI Plus at $4.99, Pro at $19.99, and Ultra from $99.99 come in. This guide breaks down exactly what “free” includes in 2026, which model you are actually talking to, how the limits work, and the specific cases people keep asking about, from students to Google Workspace to the Gemini API.

The Key Takeaways

• Gemini is free. The free plan is $0 with no card and no trial expiry.
• Since October 9, 2026 the free tier runs on Auto: most prompts go to Gemini Flash-Lite, harder ones may go to Flash or Pro, and only paid plans let you pick a model. Gemini 4 Argon starts at AI Pro.
• Free is generous but not unlimited. Usage caps refresh every five hours inside a weekly limit, video generation is paid-only, and Google publishes no numbers for either.
• Paid plans start at $4.99/month (AI Plus), then $19.99/month (AI Pro) and $99.99/month (AI Ultra).
• The free student year is back. Google relaunched it on August 19, 2026: AI Pro in the US, AI Plus in 140+ other markets, redeem by December 31.

Is Gemini free? The short answer

Yes. Anyone with a Google account can open the Gemini app on the web, iPhone, or Android and start chatting for free, forever, without paying or entering card details. This is the same product Google markets as its consumer AI assistant, not a limited free trial that flips to paid after 30 days. If you are unsure which model you are using or which is best, our Gemini model comparison breaks down all of them.

What you are getting for free is real. Free Gemini can write, code, analyse images and PDFs you upload, browse the web, generate pictures, and hold a spoken conversation. The difference between free and paid is mostly about how much you can do per day and which top-end models you can reach, not whether the core assistant works. If you want the full money breakdown, our Gemini pricing breakdown covers every plan line by line.

What you get on the free Gemini plan

The free tier changed a lot in 2026, and until October mostly for the better. Google handed free users features that used to sit behind a subscription, and from July 21 the faster Gemini 3.6 Flash as well. The October change is the first step back: since October 9, free accounts keep the features but lose the model choice, and most prompts now land on Flash-Lite, the lightest of the three in the Gemini app.

At the core you now get Auto for chat. It answers most prompts with Gemini Flash-Lite, which Google describes as an efficient workhorse for everyday tasks like summarising and brainstorming, and passes prompts that need deeper reasoning to Flash or Pro. Turn smart model selection off and every answer comes from Flash-Lite. As on every plan, you can set its effort to low, medium, or high, and the higher settings give more thorough answers while using up more of your allowance. On top of that, free accounts can still generate and edit images with Nano Banana 2, hold hands-free voice conversations through Gemini Live, and run Deep Research reports. You also get Canvas and Gems for building custom assistants, and 15 GB of cloud storage shared across Gmail, Drive, and Photos.

That is a capable free assistant. The things you do not get for free are unlimited heavy usage, video generation, the highest Deep Research limits, and the largest context windows, all of which live on the paid tiers below. For video specifically, Google’s separate Gemini Omni Flash model handles talk-to-edit generation.

Which Gemini model is free in 2026?

Since October 9, 2026, free Gemini runs on Auto rather than one fixed model. Google’s help page on changes to Gemini model access and limits makes Auto the default without a plan: Gemini sends each prompt to the model that suits it, most prompts go to Gemini Flash-Lite, and prompts that need deeper reasoning may go to Flash or Pro. If you turn smart model selection off, every answer comes from Flash-Lite. What free users cannot do is pick Flash or Pro themselves. The page gives no version numbers for these models, and neither do we.

Until that date the free tier ran Gemini 3.6 Flash, which Google shipped on July 21, 2026 to all Gemini app users, plus what Google’s plans page calls varying access to Gemini 3.1 Pro. That plans page now lists only “Access to our Flash-Lite model” under Free, which undersells Auto, so where two Google pages disagree, the dated help page is the one that describes your account. Our full Gemini 3.6 Flash breakdown has the benchmark numbers for the model free users are losing.

Above Flash-Lite sit Flash and Pro, and the October change splits them by plan. Google AI Plus keeps Flash but loses Pro, on a date Google is emailing to each subscriber. Only AI Pro and AI Ultra keep all three models, and since October 13 both also get Gemini 4 Argon, Google’s newest public flagship. Our Gemini 3.5 review covers the model that ran the free tier until July and where it still trailed 3.1 Pro.

What you will not find on the free plan is Gemini 4 Argon. Google’s newest flagship reached the Gemini app on October 13, 2026, rolling out to Google AI Pro and Ultra subscribers only, and Google has not said whether it will reach Plus or free. Gemini 3.5 Pro, which Google announced at I/O in May 2026, has still not had a public release. And no, there is no separate “Gemini 2” you need to hunt for; the older 2.5 models have been superseded by the 3.x line in the app.

Free vs paid: Gemini plans compared

Google splits Gemini into one free tier and three paid ones, all listed on Google’s official Gemini plans page. The paid plans do not hand you secret models so much as raise your limits, add video generation, and pile on cloud storage. Those limits now run on compute rather than a prompt count, as our guide to Gemini's usage limits explains. Here is how the four consumer plans line up in 2026.

The model column is the part that moved. Until October, Google’s limits page marked Flash-Lite, Flash, and Pro as available on every tier, free included, and the paid plans were sold on limits alone: 2x, 4x, and 5x or 20x. Since October 9 the model choice itself splits by plan, so the cheapest way to pick Flash yourself is now AI Plus at $4.99, and the cheapest way to pick Pro, or to use Gemini 4 Argon, is AI Pro at $19.99.

Free ($0/month)

The free plan is still the right pick for casual use. Since October 9 you get Auto, mostly Flash-Lite, plus image generation, voice, and Deep Research, with no card required. You will feel the ceiling sooner than before, though: with no way to pick Flash or Pro, harder reasoning or coding work is the signal to look at a paid plan.

Google AI Plus ($4.99/month)

AI Plus is the cheap step up, and it got cheaper. Google cut it to $4.99/month in June 2026, down from $7.99, while keeping the same package. It doubles your usage limits over free, adds video generation with 200 Google Flow credits, folds Gemini into Gmail, Vids, and Daily Brief, and bumps storage to 400 GB. It suits people who hit the free caps but do not need heavy professional use. The package does change in October, though: Google is removing the Pro model from AI Plus on a date it emails to each subscriber, leaving Flash-Lite and Flash.

Google AI Pro ($19.99/month)

AI Pro is the mainstream paid plan and the one most power users land on. Google lists it at 4× higher usage limits than free, with Gemini 4 Argon, the Pro and Flash models, the Jules coding agent, Google Antigravity, 1,000 Flow credits, YouTube Premium Lite, and 5 TB of storage. Since October it is the cheapest plan that keeps the Pro model in the Gemini app, and since October 13 the cheapest way to use Gemini 4 Argon there. This is the tier that competes head-to-head with a ChatGPT Plus subscription.

Google AI Ultra (from $99.99/month)

Ultra is the everything tier, aimed at developers and heavy creators. It launched at $249.99/month at I/O in May 2025 and now starts at $99.99, with a $199.99 option above it. The two Ultra plans differ by usage, at 5× and 20× the AI Pro limits, with 10,000 or 25,000 Flow credits, early features like Project Genie, YouTube Premium, and cloud storage starting at 20 TB.

Is the free Gemini plan unlimited?

No, free Gemini is not unlimited. Since May 17, 2026 Google has metered compute rather than counting prompts: your allowance refreshes every five hours until you reach a weekly limit, and a complex prompt counts for more than a simple one. You can send a healthy number of messages, but not an endless stream, and Google publishes no fixed figure for the free allowance.

In practice the limits you actually notice on free are Deep Research and image generation caps, plus, since October 9, having no say over which model answers. Google does not publish a single figure for any of the caps; its own limits page only says AI Plus is 2× the standard limits, AI Pro 4×, and Ultra 5× or 20× the AI Pro limits, without ever stating the baseline. Normal back-and-forth chat on the lightest model draws the least from that allowance. If you routinely max out those ceilings, that is your signal to look at AI Plus or Pro, and our guide on how much AI costs in 2026 helps you weigh it against rivals.

That answer matters more than it used to, because a rival moved. Since the week of 10 August 2026 OpenAI has not capped text chats on the free ChatGPT plan, which retired a limit of roughly 10 messages every 5 hours and now runs GPT-6 Luna, so “unlimited” is a genuine point of difference between the two free tiers. The important caveat is that OpenAI’s promise covers text only, and file uploads, images, voice, and image generation all keep their own separate caps.

So neither free plan is unlimited across the board, they just draw the line in different places. Google meters everything through one compute allowance that refreshes every five hours inside a weekly limit, while OpenAI lets plain text flow freely and holds the line on attachments and media instead. If you mostly type questions and read answers, that difference now favours ChatGPT, and our guide to whether ChatGPT is free covers exactly what survives the change.

Is Gemini free for students?

This is where the answer has changed twice. Google’s first student promotion ran through 2025 and into early 2026, then closed on staggered regional deadlines, with the final United States window expiring on April 30, 2026. It stayed shut for about four months. Then on August 19, 2026, Google relaunched it for the 2026 to 2027 academic year, and the new terms split by region: eligible college students in the United States get 12 months of Google AI Pro free, a plan that otherwise costs $19.99/month, while students in more than 140 other markets get 12 months of Google AI Plus. Both windows close on December 31, 2026.

Two details decide whether this applies to you. Students whose 2025 AI Pro trial already expired qualify again, so burning the first free year does not lock you out of this one. And the offer is not unconditional: you verify enrollment with a valid school email address, a payment method is required at sign-up, the plan auto-renews at the full price once the year ends, and eligibility has to be re-verified every year. A handful of countries are excluded outright, Canada, Hong Kong and Macau among them. Anyone who is not eligible still gets the standard free Gemini tier at no cost, which for study, writing help, and research is often enough, and our walkthrough on the free Google AI Pro for students offer covers the claim steps and the regional fine print.

Is Gemini free with Google Workspace and Gmail?

If you pay for Google Workspace, Gemini is already included at no extra cost. Google scrapped the old paid “Gemini for Workspace” add-on and now bakes the AI into the plans directly. The level of access scales with your Workspace tier.

Business Starter is thinner than it looks. Google’s edition comparison gives Starter Gemini inside Gmail and Google Vids only, leaving Docs, Sheets, Slides, Meet, Chat, Calendar, Forms and Drive blank in that column, plus the standalone Gemini app at what Google calls standard access to features and models on a 32,000-token context window. Starter gets no published prompt count at all: its Pro and Thinking rows both read “basic access, access varies, daily limits may change frequently”. Business Standard and above switch on the side panels in Docs, Sheets, Slides and Meet, add Help me schedule and Ask Gemini in Meet, and lift the app to Pro access with a 1 million token window and the only hard numbers Google prints here, up to 25 prompts per 4 hours on Pro and up to 300 per day on Thinking. Those are Workspace figures, not consumer ones; the free tier in the Gemini app still has no published cap. Either way, the Gemini help you see in Gmail is part of what you already pay for, not a separate Gemini bill.

Is Gemini free on Pixel and other phones?

Yes. On Pixel phones Gemini is the built-in assistant and it is free to use, the same free tier you get anywhere else, with no Pixel-only charge. The same applies to the Gemini app on any Android phone or iPhone, and to Samsung Galaxy devices where Gemini powers the assistant.

Owning a specific phone does not buy you extra Gemini features on its own. What a new Pixel sometimes adds is a limited-time trial of Google AI Pro bundled with the device, but once that promo ends you are back on the standard free plan unless you subscribe.

Is the Gemini API free?

For developers, Gemini has a separate free tier through Google AI Studio, and it also needs no credit card to start. Google’s pricing table marks input and output as free of charge on the free tier for Gemini 3.8 Flash and 3.6 Flash, plus 3.5 Flash-Lite and 3.1 Flash-Lite, and for the older 2.5 Pro, 2.5 Flash and 2.5 Flash-Lite. It no longer publishes the daily request and token ceilings on that page, so check your own live quota in AI Studio rather than trusting a figure from a blog.

Two catches matter here. Google’s pricing page marks Gemini 4 Argon and Gemini 3.1 Pro Preview as “Not available” on the free tier, so the current Pro-class models are paid-only and the one Pro model you get free is the older 2.5 Pro. And the moment you enable billing on a project, the free tier disappears for that project and every call becomes billable from the first token. For light experiments the free API tier is real; for production you will pay. Paid rates for 3.6 Flash are $0.75 per million input tokens and $3.75 per million output through December 31, 2026, rising to $1.50 and $7.50 from January 1, 2027.

Why is Gemini free?

Google does not run a free Gemini tier out of generosity. The free plan is a growth engine. It gets hundreds of millions of people into Google’s AI habit, feeds Gemini into Search, Android, and Workspace where Google already dominates, and creates a smooth path to upsell the heavy users onto AI Pro or Ultra.

There is also competition, and it is intensifying. With ChatGPT, Claude, and Grok all offering free access, and OpenAI dropping text limits on its free plan in August 2026, a paywalled Gemini would simply lose users. Free is the price of staying in the race, and Google can afford it because the cheap Flash models make free usage viable at scale.

Free Gemini is great, but it is one model in one app

The honest limitation of free Gemini is not the usage cap. It is that you are locked into Google’s models and Google’s app. Gemini is excellent at some jobs and second-best at others; on coding and professional tasks, for example, a rival model often edges ahead, which is exactly why so many people now compare Gemini and ChatGPT before committing. Switching apps every time you want a different strength is friction most people never bother with.

That is the gap Fello AI fills. Fello puts Gemini alongside ChatGPT, Claude, Grok, and DeepSeek in one native app, so you can send the same prompt to several models and keep the best answer instead of guessing which one to open. It is also a creation tool, not just a chat window, generating images, slide decks, documents, PDFs, and Excel files from a prompt.

For $11.99/month in the US (prices vary by country) Fello gives you access to all of those models in one place, which works either as an alternative to a single Gemini subscription or as a complement to your free Gemini account. If your workflow spans research, writing, coding, and design, one app that covers every model beats juggling five. You can download Fello AI on the App Store and try it with the free Gemini models included.

The bottom line

Gemini is free, and the 2026 free tier is still a capable one, but it has been smaller since October 9. You get Auto, mostly Gemini Flash-Lite, plus image generation, voice, and real research tools without paying a cent. For everyday questions most people never need more than that.

You only pay when you want higher limits, video generation, or the choice of Flash, Pro or Gemini 4 Argon, and that is where AI Plus at $4.99 or AI Pro at $19.99 make sense. Start free, watch where you hit the ceiling, and upgrade only if you actually run into it. And if you find yourself wishing Gemini did everything, remember no single model does, which is the whole reason multi-model apps exist. On a Mac you can run the free tier as a proper app rather than a tab, and our Gemini desktop client guide for macOS shows how.

FAQ