Blog · AI tools · Published
How to use Gemini: entry points for the web version and mobile app, sign-in requirements, and basic operations
The web version is at gemini.google.com, and the officially listed supported browsers are Chrome, Safari, Firefox, Opera and Edgium (as written in the original); some web features work without signing in, while the Gemini app on Android devices requires a Google account, and the official page does not state this for other device types. Sign-in accounts fall into three types — personal, work and school — with the threshold resting on yourself, the Workspace edition purchased by your organization plus being 18 or older, and a switch controlled by the school administrator, respectively. When not signed in, restricted features fall into two groups: one depends on history, preferences or connected services, the other requires only account authentication. The article provides an entry-point comparison table, the differences between the three account types, five steps for a first conversation, four common actions after signing in, and an order for troubleshooting by symptom; no promise is made about available regions.
To use Gemini, open gemini.google.com in a desktop browser and you can get started; the officially listed supported browsers are Chrome, Safari, Firefox, Opera and Edgium (as written in the original; usually meaning Edge). Some features work without signing in, but to use more features or save activity, you need to sign in with a Google account (Google's official help page). The Gemini app on Android phones requires sign-in; for iOS and other device types, this page does not give the same conclusion, and this article goes only as far as the official documentation.
1. Web version: one entry point, one browser list
The web version has only one entry point: open gemini.google.com in a desktop browser. The official list of usable browsers has five items — Chrome, Safari, Firefox, Opera and Edgium. The last item is written as it appears in the original page.
This looks trivial, but it is the first thing to rule out when troubleshooting: if the page opens but the input box cannot be clicked, or the layout is clearly broken, switch to any of the browsers above first and look again. Suspecting the account or the network at this point usually means looking in the wrong direction. The official documentation also provides separate pages on the languages, countries and regions supported by the web version and the mobile app; these two matters are governed by the official pages, and this article does not reproduce the lists, nor does it recommend bypassing regional restrictions in any way.
2. What works without signing in, and what you lose
The official documentation states that some features can be used without signing in, and that what is available varies by device type and region. Without signing in, you can generally do text-based interactions, such as asking an informational question or having it generate a short piece of text; these capabilities may also be adjusted.
What you genuinely cannot get without signing in falls into two groups, for different reasons.
The first group relates to data in the account: you cannot see past conversations; answers are not personalized based on your previous conversations or preferences; connected apps (Connected Apps, which bring in services within your account such as Google Workspace and Google Photos) are unavailable.
The second group is listed by the official documentation directly under features requiring account authentication: image generation, Gems (custom assistants that save frequently used instructions as dedicated assistants), and file uploads.
The point of grouping them is that troubleshooting goes in different directions. If image generation errors out or uploads do nothing, check sign-in status first; if you cannot find yesterday's conversation, or the answer does not seem to continue from last time, that is a problem on the history and account side. Mixing the two groups together tends to mean repeated attempts in the wrong direction.
3. The three account types that can sign in, with the threshold held by different parties
The official documentation lists three Google account types that can sign in:
- A personal Google account.
- A work Google account: it must be an eligible Workspace edition (Google's office suite, including Gmail, Docs and so on), and the user must also be 18 or older.
- A school Google account: it can be used only after the school administrator enables the service.
The difference between the three is not whether you have a Google account, but who decides whether you can get in: for a personal account, you decide yourself; for a work account, it hinges on the edition the organization purchased, plus the age requirement; for a school account, it depends only on whether the administrator has enabled it. If you cannot sign in, first confirm which type you have, then check against that threshold, and do not apply the work account's two conditions to a school account for an extra round of checking.
When an account is managed with Family Link (Google's family management tool), the administrator can turn off access to the Gemini app for users under 13, or under the applicable age in their country.
4. Web version, Android app and temporary chats: different entry points, different sign-in requirements
| Entry point | How to access | Sign-in required | What works without signing in |
|---|---|---|---|
| Web version (desktop browser) | Open gemini.google.com in a browser | Some features work without signing in | Generally only text-based interactions |
| Gemini app on Android | Install and open the app | Must sign in with a Google account | Not usable |
| Temporary chat | Within the web or mobile app, enter next to New chat | Sign-in required, personal accounts only | Not usable |
| Mobile app on other device types | See the corresponding official documentation page | Not stated on that page | Subject to the official documentation |
The third row of the table deserves separate mention. A temporary chat is not a separate product, but a way of chatting within the web or mobile app: it does not go into activity, nor into recent conversations, and once you leave you no longer have a chance to open it again, and a temporary chat indicator stays on screen the whole time.
The official documentation is clear about the trade-offs: Gems are unavailable; connected apps that depend on the activity setting being retained are unavailable; there are no personalized replies; information is not saved for future personalization; and you cannot give feedback on answers. It is available only to personal Google accounts, and cannot be used by work or school accounts, and Gemini in Chrome and Gemini in Google Messages do not have this entry point either. So it suits one-off questions where you do not want to leave a trace, and is not suitable as a daily mainstay.
5. Your first conversation on the web version: five steps
- Open gemini.google.com on a computer.
- Enter your question or prompt in the input box at the bottom of the page.
- If you need to include a file or image, click Add files next to the input box and add it following the on-screen instructions.
- Click Submit.
- To start a new conversation, click New chat in the top left; if you cannot see this button, click Menu first.
For the first time, pick a task whose result is easy to verify. The official examples include: outlining a blog post, explaining an error message, creating a study plan with recommended videos, turning a set of expense data into a pie chart, and writing a caption for a photo. Using the one for which you happen to have the material on hand makes it easier to judge whether the answer is right than thinking one up out of thin air.
The input box is at the bottom of the page; if you do not see it, first confirm the page has finished loading — there is no need to dig through the menu.
6. Four actions worth learning first after signing in
Edit the prompt and regenerate. Move the mouse to the right side of your prompt, click Edit text to edit the text, then click the update button on the page; Gemini regenerates according to the new prompt, without rewriting the whole thing.
Choose the model manually. By default the system automatically picks a suitable available model based on your prompt; to choose yourself, you must sign in first, then select from the menu.
Turn results into files. You can directly ask for the answer to be generated as a file; the officially listed formats include Workspace files (Docs and Sheets), .pdf, .docx, .xlsx, .csv, LaTeX, plain text TXT, RTF and Markdown; most formats can be downloaded directly to your device, or exported to Google Drive.
Reuse content from the answer. There is a copy button below code blocks; if you want a table in the answer turned into a chart, simply ask it in a follow-up to draw a chart with that data; content such as itineraries, training plans and to-do lists appears as interactive cards, which may include an add-to-calendar button, and you can also ask it in the conversation to change the card content directly.
7. When a feature does not work, check in this order
| Symptom | Check first | Why |
|---|---|---|
| Image generation or file upload errors | Sign-in status | Both are listed by the official documentation under features requiring account authentication |
| Buttons unresponsive, layout broken | Browser | The web version is supported only on the browsers listed officially |
| Cannot sign in with a work account | The organization's Workspace edition and age | The official documentation states both the edition and the age of 18 for work accounts |
| Cannot use a school account | Whether the administrator has enabled it | A school account can only be used after the institution's administrator enables it |
| All of the above are fine but it still will not open | The official language and region documentation pages | Availability is described by the official pages, and this article does not make a judgment |
The point of the order is to save back-and-forth trial. The first four can all be self-checked within a few minutes; if it still does not work after that, then check against what the official documentation pages show at that time, rather than suspecting the service itself is broken first.
8. Two official notices that are easy to overlook
Gemini can make mistakes. The official documentation asks users to verify answers themselves and not to treat them as professional advice; for content such as amounts, health and legal provisions, verification should be part of the process, not an option.
Gemini may use your approximate location, or, when you allow it, the device's precise location, to make answers more relevant. On the web version, the path to view this is: open gemini.google.com, click Menu at the top to go to Settings, and look at the small dot next to the location indicator at the bottom of the menu: blue means precise location is in use, gray means it is not; click Update location to grant permission. Location management in the mobile app has a separate documentation page.
9. This site's relationship with Gemini
This site provides third-party top-ups for ChatGPT memberships; it does not sell Gemini accounts, nor does it sell Google AI plans. The terms of service state that this site is a third-party top-up provider and that membership features and quotas are subject to OpenAI's official rules at the time; the terms themselves address ChatGPT top-ups and do not cover Gemini or Google services (Top-up service terms); the product page likewise lists ChatGPT membership top-up tiers, with no Gemini-related products (ChatGPT top-up plans).
If what you want to compare is how to choose a tier on the ChatGPT side, ChatGPT subscription cost overview gives a framework by use case; for Gemini's own usage and upgrade differences, see Gemini usage limits and upgrades.