Claude vs ChatGPT vs Gemini — A Real-World User’s Honest Take

Last updated: March 30, 2026

I don’t often review tech — this is only my third post on it — but I use AI tools a lot. I promote my own work, and I can’t afford to pay someone else to do it for me. That means I need the help that I can get. And just so we’re clear — my books are not AI generated. I use AI to help edit and promote them.

I figure I get about a month’s worth of work from these tools for what I’d pay a person for a single hour. That’s a pretty good deal.

Over the past couple of years I’ve put Claude, ChatGPT, and Gemini through their paces — not for testing purposes, but because I had real work to do. Here’s what I’ve found.

First, two things that matter more than which one you pick

Before anything else: if you’re doing anything serious with AI, make sure you’re using the latest model version. For serious ongoing work, the paid plans are where these tools become much more useful — don’t judge any of them by the free tier.

The other thing worth knowing: all three now offer some form of continuity across chats, but they handle it differently. ChatGPT can reference past chats when that setting is on. Claude has memory plus a past-chat search feature on paid plans. Gemini offers memory of past chats for eligible users, and Google says that’s gradually expanding. The frustrating blank-slate experience has mostly gone away, but how well it works varies.

One more hard-learned tip: if you’re using AI for feedback on your writing, ask for the top changes only, and give it one chapter at a time. Early on I’d ask for feedback and it would change about every fifth word — basically rewriting my story. Limiting what I ask for made all the difference.

Claude — best for matching your voice

If you create content that needs to sound like you, Claude is where I land. It picks up on tone and style faster than the others, and it requires fewer changes to be something I’m comfortable with. For drafting emails, editing content, or handling promotional work — it’s the strongest of the three.

Worth knowing: Claude currently isn’t a photo or illustration generator the way the other two are. It can analyze images and produce diagrams or interactive visuals, but if you need actual generated images, you’ll be going elsewhere. For text work though, it consistently gets closest to what I actually wanted.

ChatGPT — best for image generation, but watch the voice

In my experience, ChatGPT produces the best still images of the three. If you need visuals for social media, blog headers, or promotional graphics, it’s my go-to.

The catch is the text. It tends toward wordy, formal writing with a lot of big words — and if you’re trying to sound like yourself, you have to push back more. It’s not bad, it just has a style it gravitates toward.

Gemini — fastest for images and video, and better than I expected

Gemini now has memory of past chats for eligible users, which is a big deal — honestly, that’s why I didn’t use it as much before. With that in place, it’s a lot more useful for ongoing work.

In my experience, Gemini is the quickest of the three when I want image or short video output fast. The images don’t always match exactly what I asked for, but it’s improving. The video generation is genuinely impressive for speed.

What surprised me recently: when the other tools were changing my words more than I wanted, I tried Gemini and found it more hands-off with my text. It’s a little less creative, and I found that worked well for what I needed at the time.

The thing nobody tells you — use all three

I don’t think of these as competitors where you pick one and commit. I think of them as a team of co-workers with different skills, in different departments.

For one recent project, I bounced ideas between all three — developing concepts with one, refining language with another, generating visuals with the third. The result was better than anything I could have done with just one of them, or on my own. Each one pushed the thinking in a slightly different direction, and that friction produced something better.

The bottom line

They all have strengths and weaknesses, and those strengths and weaknesses are a moving target — they update constantly. The best approach is to know what each one is good at right now, and match the tool to the task.

For my work, Claude handles the words, ChatGPT handles the images, and Gemini keeps things grounded when I need something done without too much reinterpretation. Your experience may be different — but that’s mine.


Tony B. Richard is the co-author of the Earth’s Secret Alliance sci-fi series. Find the books at tonybrichard.com/books.