So a friend texted me last week asking "wait, what even is Gemini, is it like ChatGPT?" and honestly, that's a fair question. Google's AI has changed a lot this year, and if you haven't opened the app since 2024, you'd barely recognize it. Here's what it actually is and how to start using it without wasting time.
Okay But What Is Gemini AI
Gemini is Google's AI assistant, built into an app, into Search, into Gmail, Docs, Sheets, basically every corner of Google's ecosystem. Think of it as Google's answer to ChatGPT, except it's got a home-field advantage: it already lives inside the apps you use every day.
To be fair, that's the whole pitch. You're not switching tabs to ask it something. It's already sitting in your inbox, your spreadsheet, your phone's assistant button.
Under the hood, Gemini isn't one single model anymore. As of mid-2026, there's a small family:
- Gemini 3.5 Pro – the heavyweight, best for complicated reasoning, long documents, and mixing text, images, audio, and video together
- Gemini 3.5 Flash – the fast, cheap version most people actually interact with day to day, and it now powers the "Auto" mode in the app
- Gemini Spark – a background agent that can handle tasks on its own, like scheduling or managing your inbox, even while your phone is locked
- Gemini Omni – a newer model focused on generating video from a mix of text, image, and audio prompts
If you've used Claude or ChatGPT before, Gemini 3.5 Pro will feel familiar as a chat-based assistant. The bigger difference is how deep it's woven into Google's other products.
What Actually Changed Recently
Google rolled a lot of this out at its I/O event back in May 2026, and it wasn't small stuff. The app got a full redesign, called "Neural Expressive," with a new pill-shaped prompt box and answers that show key info up top instead of dumping a wall of text on you.
They also switched how usage limits work. Instead of counting a fixed number of daily prompts, Gemini now measures "compute used," meaning a quick text question costs less than, say, asking it to analyze a video or write code. Limits refresh every five hours until you hit your weekly cap, according to Google's own announcement.
Then in June, Google added Gemini 3.5 Live Translate, an audio model that handles real-time speech translation across more than 70 languages while keeping your actual tone of voice. That one's genuinely useful if you're on multilingual calls or traveling a lot.
How To Actually Use It
Honestly, getting started is the easy part.
- Go to the Gemini app or website and sign in with a Google account
- Type a question like you're texting a smart friend, no need for weird prompt formulas
- Use the "+" menu to attach files, images, or turn on specific tools like Deep Research
- If you're a Workspace user, you'll also see Gemini show up as a side panel in Docs, Sheets, Gmail, and Meet without installing anything extra
For everyday stuff, the free tier covers a lot. If you want the heavier features, longer chats, or more image and video generation, Google offers paid tiers: AI Plus, AI Pro, and now AI Ultra, which starts at $100 a month after a recent price restructure.
Why This Matters
Here's the thing that stands out to me: Google isn't trying to make Gemini feel like a separate AI product you have to remember to open. It's trying to make it invisible, just a layer sitting under stuff you already do. That's a different bet than what OpenAI or Anthropic are making with standalone assistants.
Whether that's better depends on you. If your life already runs through Gmail and Docs, Gemini's integration is genuinely convenient. If you want a more focused chat assistant without the ecosystem baggage, you might still prefer something like ChatGPT or Claude.
What's Next
Google has said Gemini 3.5 Pro was still rolling out more broadly as of the I/O announcement, so not every feature mentioned here may be live in your account yet. Gemini Spark was also described as being in testing at launch, so treat anything agent-related as still maturing rather than fully proven out.
Have you tried the new Gemini app yet, or are you still team ChatGPT? Curious what's actually working for people right now.
){kind=link}