- Go to gemini.google.com in any browser, or install the Gemini app on Android or iPhone. Sign in with a Google account and start typing.
- The free plan is enough for most people: it runs on Gemini 3.6 Flash, gives limited access to the smarter 3.1 Pro model, and includes image generation, Deep Research, Gemini Live voice chat and Canvas.
- Type or say what you want in plain English. No special commands. Attach a photo or PDF with the paperclip icon.
- Turn off Gemini Apps Activity in your account settings if you don’t want your chats reviewed to improve Google’s AI.
Google Gemini is Google’s AI chatbot — the direct competitor to ChatGPT and Claude. This guide shows you where to find it, what you get without paying, and how to use it for real, everyday tasks: writing an email, summarising a PDF, making an image, or just talking to it out loud.
Where to find Gemini
You don’t need to download anything to try Gemini on a computer. Open a browser and go to gemini.google.com. Sign in with any Google account (the same one you use for Gmail is fine) and you’re in.
On a phone, install the Gemini app from the Google Play Store on Android or the App Store on iPhone. On many newer Android phones Gemini has already replaced Google Assistant, so long-pressing the power button or saying “Hey Google” opens it.
You’ll also see Gemini turn up inside Gmail, Google Docs, Sheets, Slides and Meet as a small sparkle icon — that’s the same assistant, built into the app you’re using.
What the free plan includes
Most beginners never need to pay. According to Google’s own subscriptions page, the free plan gives you:
- Unlimited chat with Gemini 3.6 Flash, Google’s fast everyday model
- Limited access to Gemini 3.1 Pro, the smarter model for harder questions
- Image generation, Deep Research (long, sourced reports), Gemini Live (voice conversation) and Canvas (a side-by-side writing and coding workspace)
- A 32,000-token context window — roughly 24,000 words of text it can “see” at once in a chat
- 15 GB of Google storage shared across Drive, Gmail and Photos
A quick word on jargon: a token is a chunk of text roughly the size of a short word. The context window is how much of your conversation and any attached files the AI can hold in mind at one time. And a model is the underlying brain — Google has several, and Flash is faster while Pro is slower but more careful.
The free limits reset every 5 hours, up to a weekly cap. If you hit a limit, Gemini tells you and usually offers to switch to the lighter Flash model so you can keep working.
Paid plans at a glance
You only need a paid plan if you use Gemini heavily or want video generation and the 1-million-token context window. Prices below are US monthly, from Google’s subscriptions page:
| Plan | Price (US) | What you get on top of Free |
|---|---|---|
| Free | $0 | 3.6 Flash, some 3.1 Pro, 32K context, 15 GB storage |
| Google AI Plus | $4.99/month | 2× the free limits, 128K context, video generation, 400 GB storage |
| Google AI Pro | $19.99/month | 4× the free limits, full 3.1 Pro access, 1M-token context, 5 TB storage |
| Google AI Ultra | $99.99 or $199.99/month | Highest limits, earliest features (two tiers offered) |
Google also gives eligible college students a free Google AI Pro plan for one year — worth checking if that’s you. For a full breakdown of what’s actually different between the tiers, see our Gemini plans guide.
How to actually talk to Gemini
There is no secret prompt language. Type what you want the way you would ask a smart, patient colleague. The better you describe the situation, the better the answer.
A useful pattern is: role + task + context + format. For example: “You’re a helpful tutor. Explain photosynthesis to a 12-year-old in about 150 words, with one everyday example.”
Everyday task 1: rewrite an email
Paste your draft and say: “Rewrite this email so it sounds friendly but professional. Keep it under 120 words. Don’t add anything I didn’t say.” Then read it carefully before sending — Gemini can invent details if you don’t tell it not to.
Everyday task 2: summarise a PDF or photo
Click the paperclip or plus icon, upload a PDF, a photo of a document, or even a screenshot. Then ask: “Summarise this in 5 bullet points and list any dates or amounts.” Great for bills, school forms and long articles.
Everyday task 3: make an image
Just describe it. “A watercolour illustration of a ginger cat asleep on a stack of old books, warm lamplight, cosy mood.” Gemini will produce a few options. If a face or hand looks off, ask it to redo that part. If you’re new to this, our beginner’s guide to AI images walks through prompt wording in more detail, and there’s also a roundup of free AI image generators if you want to compare.
Everyday task 4: Deep Research
Look for the Deep Research option in the prompt bar. Ask something like: “Compare the top three robot vacuums under $400 in 2026. Cover suction, pet hair, app quality and warranty. Cite sources.” Gemini will plan the research, browse the web for several minutes, and hand back a report with links. Always click through the links — AI research tools sometimes summarise a source in a way that shifts its meaning. If you like this style of tool, compare it with Perplexity Pro and NotebookLM.
Everyday task 5: Gemini Live (voice)
Tap the waveform icon in the mobile app or on the web. You can now talk to Gemini and it will talk back, interrupting and reacting naturally. It’s genuinely useful for practising a language, rehearsing an interview, or thinking out loud on a walk. On phones with a camera share option, you can also point the camera at something and ask about it.
Everyday task 6: Canvas
Open Canvas when you’re writing something longer — a cover letter, a school essay outline, or a bit of code. Your document sits on one side, the chat on the other, and you can ask Gemini to edit specific paragraphs without rewriting the whole thing. Please still write in your own voice where the work is meant to be yours; passing off AI text as your own is against most school and workplace rules.
Privacy: what Google keeps
By default, Google saves your Gemini chats and may have human reviewers read a sample to improve the AI. You can change this. In the Gemini app or on the web, open your account menu and look for Gemini Apps Activity. You can pause it, set chats to auto-delete after a period, or delete individual conversations.
Two practical rules: don’t paste passwords, ID numbers, or medical details you wouldn’t put in an email; and treat Gemini’s answers as a well-read friend’s opinion, not a doctor, lawyer or accountant.
When Gemini gets it wrong
All chatbots occasionally invent facts — names, dates, quotes, court cases. This is called hallucination. Reduce it by:
- Asking Gemini to cite sources and then actually opening them
- Uploading the source document yourself instead of trusting its memory
- Using Deep Research or the 3.1 Pro model for anything that matters
- Asking a second AI. Our Claude vs ChatGPT comparison covers the two obvious alternatives, and this list of ChatGPT alternatives covers the rest.
Gemini vs ChatGPT: which should a beginner pick?
Both are excellent and both have generous free tiers. Gemini has the edge if you live in Gmail, Docs and Android; ChatGPT has the edge if you want the biggest ecosystem of custom “GPTs” and voice quality. If you’re weighing the paid tiers, our guide to ChatGPT Free vs Go vs Plus vs Pro pairs well with the Gemini plans link above.
Frequently asked questions
Do I need to pay to use Gemini?
No. The free plan covers Gemini 3.6 Flash with unlimited chat, some access to the smarter 3.1 Pro model, image generation, Deep Research, Gemini Live and Canvas. Most people never need more. Paid plans mainly add higher limits, video generation, a bigger context window and more Google storage.
What’s the difference between Gemini Flash and Gemini Pro?
Flash is fast and lightweight — great for everyday questions, rewrites and short summaries. Pro is slower but reasons more carefully, so it’s better for complex research, tricky maths and longer documents. On the free plan you get unlimited Flash and a smaller allowance of Pro.
Can Gemini see images and PDFs I upload?
Yes. Use the paperclip or plus icon to attach a photo, screenshot or PDF, then ask questions about it — “what does this error message mean”, “summarise this contract”, “what plant is this”. Don’t upload documents containing sensitive personal information you wouldn’t email.
Is my Gemini conversation private?
By default Google stores chats and may sample them for human review to improve the product. You can turn this off under Gemini Apps Activity in your Google account, and delete past chats. Even so, treat the chatbot like any online service: don’t share passwords, ID numbers or confidential work.
Can Gemini make videos?
Video generation is not included on the free plan — it starts on Google AI Plus and above, per Google’s subscriptions page. If you just want to try AI video without paying, see our roundup of the best free AI video generators.
Is Gemini safe for kids and students?
Google offers a version of Gemini for supervised child accounts through Family Link, with stricter safety settings. For older students, the assistant is useful for explanations and study help, but check your school’s rules — many require you to write your own final answers rather than submit AI-generated work.