Gemini

Google's AI assistant. Near-instant answers, good at reading images and video, excellent search, and simple Gems that turn it into small task agents. My go-to for quick, small jobs.

Free / $4.99/month Google AI Plus / $19.99/month Google AI Pro / from $99.99/month Google AI Ultra

My take

Gemini isn't one of the strongest AI tools, and for real work I use ChatGPT Codex much more. But it's still my go-to for tiny work packages I need done quickly, like translating UI copy into other languages or getting UX feedback. The biggest plus for me is the near-instant answers. It also reads images and videos well, and its search is excellent. My Gems are tiny agents with just enough context and instructions for those few tasks. They're very basic compared to what others have built, but that's the charm: like the early custom GPTs, they're easy to maintain. Four stars, for Gemini Flash and the image and video models that come with it.

Visit Gemini Back to tools

At a Glance

Pros

+ Near-instant answers with Gemini Flash
+ Reads images and videos well
+ Excellent search
+ Gems are simple task agents that are easy to maintain
+ Google's image and video models built in
+ Built into Google Workspace

Cons

− Not one of the strongest tools for bigger, demanding work
− Gems are basic compared to what other tools offer, and Google is replacing them with Skills from October 2026
− Confusing names and plans (Bard → Gemini → Google AI Plus, Pro and Ultra)

Best for: Small jobs you need done fast: translating UI copy, UX feedback on screenshots, quick lookups, reading images and video

Powered by

Gemini 3.8 Flash multimodal Gemini 4 Argon multimodal Nano Banana 2.1 image

Full Review

Gemini isn’t the strongest AI tool I use. For real work, I reach for ChatGPT Codex much more often. But Gemini has one thing going for it that I value a lot: it answers almost instantly. For a small job I need done right now, that speed matters more than extra depth.

What I use it for

Small, well-defined jobs:

Translating UI copy. Button labels, error messages, short onboarding texts into other languages.
UX feedback. Quick reactions to a screen or a flow. Because it reads images well, it can react to what’s actually there.
Quick lookups. Its search is excellent, so a fast question with a current answer is a good fit.

Anything bigger goes to a stronger tool.

Gems: tiny agents, easy to maintain

I don’t use Gemini as a sparring partner. I use a few Gems: custom versions of Gemini with a short set of instructions and just enough context for one task. Compared to what other tools let you build, Gems are very basic, and that’s the charm. They’re like the early custom GPTs: a name, some instructions, maybe a file or two. Setting one up takes minutes, and keeping it current is just as quick.

Gems are becoming Skills

Google has announced that Skills will replace Gems. Skills start rolling out in the Gemini app on 13 October 2026, and from 17 November Gems move to the settings panel. Existing Gems carry over, including their instructions and knowledge files. A Skill is a reusable set of instructions you call with a slash in a chat, and you can combine several. On paper that keeps what I like about Gems. Whether it stays as simple to maintain is what I’ll be watching.

Images and video

Gemini comes with Google’s own image and video models. Nano Banana makes the images (Nano Banana 2.1 since 6 October 2026, replacing Nano Banana 2). Videos in the app now come from Gemini Omni Flash, which Google says replaces Veo there.

Google Workspace

If your company runs on Google Workspace, Gemini may well be the AI you’re given at work. It sits directly in Gmail, Docs and Vids.

Who should use Gemini

Anyone who wants fast answers for small jobs, works a lot with screenshots, images or video, or spends the day in Google Workspace. The free tier is enough to find out whether the speed suits you. Gemini 4 Argon arrived in the Gemini app on 13 October 2026 for Google AI Pro and Ultra. For bigger, demanding work, pair it with a stronger tool.