Gemini is Google's AI assistant, which replaced Bard and Google Assistant and is now built into Google Search, Android, Gmail and other Google services. You can chat with it, send it documents and video, generate images and hand it multi-step tasks. Its main difference from competitors is its direct link to Google's world: if your email, files and calendar live in Google, Gemini can reach them without copy-pasting.

What is Gemini good for?#

  • Working with Google email and files: You ask it to find and summarize a client's emails from last week, or to build a report from a Drive file.
  • Reading long documents: Its large context window lets you hand over a whole thesis or several contracts at once and ask questions about them.
  • Summarizing YouTube videos: A student pastes the link to a long lecture and gets a list of key points.
  • Generating and editing images: For a shop post or featured image, it creates a picture or edits an existing photo from a text instruction.
  • Deep research: For researching a product market, it reads dozens of sources and returns a tidy, sourced report.
  • Android phone assistant: On the phone it works by voice and can see and explain what is on the screen.

Strengths and weaknesses#

Strengths:

  • Deep integration with Gmail, Drive, Calendar, Maps and YouTube
  • Very large context window for long documents and video
  • Strong multimodal features across image, audio and video
  • A generous free version with any Google account

Weaknesses:

  • Sometimes weaker than competitors at following complex instructions and output formats exactly
  • Product and plan names change constantly, which makes it confusing to follow
  • Some features are active only in certain countries or for paid accounts
  • Your data stays in Google's ecosystem, which is a privacy concern for some users

Free version and pricing#

Gemini is freemium. The free version gives limited chat, file uploads and image generation with any Google account. Google's paid plans add stronger models, higher limits, video generation, more storage and Gemini inside Gmail and Docs. The API also has a free quota for developers and pay-as-you-go pricing. Check the official site for current prices.

Language support#

Gemini's multilingual ability is good: it understands casual and formal text, writes fluently and offers voice conversation in many languages. For specialist translation and technical terms, as with any model, proofreading is needed.

Getting started in three steps#

  1. Sign in to the Gemini site or app with your Google account.
  2. In settings, turn on the Gmail and Drive connections only if you need them.
  3. For a long document, upload the file and ask a specific question, not a generic "summarize".

Gemini alternatives#

  • ChatGPT: more all-round, with more learning resources.
  • Claude: stronger at natural writing and following instructions precisely.
  • Microsoft Copilot: plays the same role for Windows and Office users.
  • Perplexity: more accurate if what you mostly want is search with sources.

To choose between the main assistants, read ChatGPT vs Claude vs Gemini: which for what. To make use of Gemini's large context window, working with long documents using AI: summarize, ask and compare and tokens, context windows and hallucination: three concepts to know before using AI will help.