Learning how to use Gemini for work pays off most when your day runs in Gmail, Google Docs, Sheets and Drive, because Gemini works inside the document itself there. This Gemini guide explains how to get started, which plan and model to choose, what the features do and what to watch for with work data.
What Google Gemini is and how the underlying model works is covered in a separate article; this guide is about everyday use. In the overview of AI models, Google’s Gemini appears alongside ChatGPT, Claude and Grok, among others.
When Gemini fits your work
Gemini fits your work when your day runs largely in Gmail, Google Docs, Sheets and Drive. Gemini works inside the document there, so you do not have to copy text into a separate window.
Gemini is a less obvious choice for teams that work in Microsoft 365 or want an assistant that works independently of any one ecosystem. That connection with your work environment is also the biggest difference between Gemini and ChatGPT in practice.
How to use Gemini: getting started
To use Gemini, sign in with a Google account at gemini.google.com or in the Gemini app for Android and iOS, then type or speak your question. Two choices then determine what you get: the plan and the model.
Choosing a plan
Your Gemini plan mainly determines the usage limit, storage and the integration with Gmail and Docs. Gemini in Gmail requires at least Google AI Plus and Gemini in Docs requires Google AI Pro, while the Gemini plans also differ in model and context window.
Choosing a model
Which Gemini model you get follows from your plan: Flash-Lite in the free version, Flash with Google AI Plus, and Gemini 4 Argon, Pro and Flash with Google AI Pro and Ultra. Choose Flash for short, everyday tasks and Pro or Argon for long documents and multi-step analyses.
Gemini’s features for work
For professional use, Gemini consists of four parts that you use separately.
Gemini in Google Workspace
Gemini in Google Workspace writes and summarises in Gmail and Docs, creates formulas in Sheets and searches your files in Drive. Ask what was agreed with a client last month, and Gemini answers from your own emails and documents.
Deep Research
Deep Research in Gemini searches a large number of web sources on its own and delivers a report with citations. The feature is included in the free version, although at busy times Google limits compute-intensive features for users without a subscription first.
Gemini Live
Gemini Live is the voice feature that lets you hold a conversation in everyday language, interrupt and ask follow-up questions. Gemini Live also works in languages other than English, including Dutch, on Android and iOS.
Gems and skills
Gems are fixed instructions you set up once and then reuse, for example for a set writing style or a recurring check. Google is replacing Gems with skills, which run automatically as soon as a prompt matches the task, and is converting existing Gems automatically.
What to watch for with work data in Gemini
With work data, the type of Google account determines what happens to your input. With a personal account, even on a paid plan, Google uses conversations to improve AI models as long as saving your Gemini activity is switched on. Human reviewers also assess some of those conversations.
With a Workspace account, Google does not use chats and uploaded files to train models. Client personal data therefore belongs only in a business account, and output only goes out after a check against the source.
Building a way of working with Gemini
A way of working with Gemini starts with one recurring task from your own work, not with all the features at once.
- Choose a task you do every week, such as summarising a long email thread or rewriting a document.
- Save the instruction in a Gem or skill, so the output is produced the same way every time.
- Refer to your own files in the prompt, for example: “Summarise this month’s emails with this client in five points and give the agreement and the date for each point.”
- Check figures, names and sources before the output leaves your organisation.
Learning more about Gemini
You learn more about Gemini one feature at a time, with the articles in this guide setting out the steps and choices. Besides Gemini, Google also releases Gemma, an open model you run on your own hardware, separate from the Gemini app.
For teams that want to use Gemini responsibly, the AI Literacy Course covers which data you can enter into an AI tool, in light of the GDPR and duties of confidentiality. It is vendor-neutral and ends with a certificate.
