Google Gemini

What Gemini is

Google runs a four-tier ladder: Free, AI Plus, AI Pro ($19.99), and AI Ultra — which itself split into two levels this year, $99.99 for 5x Pro’s usage and $199.99 for 20x, after Google halved the old $249.99 Ultra price at I/O 2026. All four paid tiers now include Deep Research, Gems, Canvas, and Gemini Live voice mode; the real differences are usage headroom and access to the sharper reasoning model, not which features you can see on the page. Free runs Gemini 3.6 Flash by default, with a daily slice of the stronger Gemini 3.1 Pro when a question actually needs it.

Where it’s genuinely good

The thing that separates Gemini from a chatbot with a research mode bolted on is how much of it can actually touch your own material. Ask Deep Research a question and it defaults to Google Search, but you can also point it at your Gmail, your Drive files, or a NotebookLM notebook you’ve already built — which means the research workflow can start from documents you already have instead of only the open web. Canvas carries that further: what starts as a conversation can end as an actual document, slide deck, or small working app, not just text you copy elsewhere.

Image and video generation are where Gemini pulls ahead of Claude specifically — Nano Banana Pro handles image work natively, in the same subscription, at a point where Claude has nothing comparable at all. Jules, still in beta, adds a genuine coding-agent option for people who want that without opening a separate tool.

Where it falls short

Google’s AI surface is genuinely fragmented. The Gemini app, Workspace’s built-in AI, NotebookLM, and AI Studio all do different but overlapping things, and it’s not always obvious which one you actually need for a given task — that’s a real cost in time, not just a minor annoyance. The free tier’s research cap (five Deep Research reports a month) will feel tight fast if research is a regular part of what you do. And the pricing ladder itself has been anything but stable this year: Google cut Ultra’s price in half and split it into two tiers in May, which is a good sign for new subscribers but makes any pricing comparison here a snapshot, not a fixed fact.

Which plan to actually get

Start on Free — it’s a real product, not a stripped preview, and the daily allotment of the stronger reasoning model means you’re not stuck with the lightweight one for everything. AI Plus exists mostly as a cheap step up if Free’s limits bother you but $20 feels like too much commitment yet. AI Pro at $19.99 is where most people should land if Gemini becomes a daily tool — it’s priced to match ChatGPT Plus and Claude Pro for a reason. Only go to Ultra if you’re actually hitting Pro’s limits regularly; with two Ultra tiers now, check whether 5x or 20x the usage is what you need before picking the more expensive one by default.

Ease of use

The core chat is as easy to pick up as anything else in this category — type, refine, done. What takes longer to learn is the ecosystem around it: which capabilities live in the Gemini app itself versus inside Gmail, Docs, or a separate product like NotebookLM. For someone who already lives in Google’s tools daily, that friction mostly disappears because the surfaces already feel familiar. For someone coming in cold, it’s the one part of Gemini that takes real orientation.

Who Gemini’s actually for

Anyone already paying for Google One or Workspace, or whose actual working materials — email, documents, calendar — already sit inside Google’s ecosystem. It’s also the clear pick if native image and video generation matters to you and you don’t want a second subscription just for that. If you’re not already in Google’s world and don’t need visual generation, ChatGPT or Claude are simpler products to reason about on their own.

xperiences.