I’ll be honest: of everything that goes into shipping an app, design is the part that eats the most of my time. I’ve been writing code for over a decade, and I can get a screen looking the way I want. It just takes a while to get there. Spacing, color, hierarchy, that endless “why does this feel off” loop. I’d happily lose an afternoon nudging pixels until it finally clicked.

Lately that’s changed. Claude Design quietly turned into my design partner, and I don’t think I’m going back.

The tool I almost paid for

Before this, my crutch was impeccable.style. It got me somewhere decent when I had no idea where to start. I was also this close to trying ui.sh. It was sitting in an open tab, basically one click away from a subscription. Then a few days later Claude Design showed up. Lucky timing, honestly: I never ended up subscribing to ui.sh.

Now the flow is simple. I describe what I want in plain words, something like “a settings page with a profile header, a list of toggles, and a danger zone”, and I get back real, working UI. Not a mockup I have to rebuild, but something I can drop straight into the project, with sensible spacing and a layout that already makes sense. Then I just keep talking to it: “make the danger zone feel more serious”, “tighten the vertical rhythm”, “this looks too generic, give it more personality.”

It’s the iteration that gets me. What used to be an afternoon of fiddling is now a five-minute conversation.

Mobile: from Expo to Flutter

Mobile was its own little adventure. I started on Expo (React Native) and gave it a real shot across a few attempts. Eventually I tried Flutter instead, and honestly, I was impressed. What Claude Design handed me translated almost directly into the Flutter result. No wrestling to make the screen match the idea; it just came out the way it looked in the design.

I might circle back to Expo and React Native somewhere down the line. But for now, Flutter plus Claude is comfortable enough that I don’t feel the itch to switch.

It’s not just prototypes

The fun part is that some of this is already live, not just sitting in a design tab. holonesia.com shipped straight from Claude Design output, and on mobile there’s Qalbi up on the Play Store.

There are a few more I built the same way. Those never turned into real apps though; they were more like playgrounds where I poked around inside Claude Design to see how far I could push an idea. Here’s one of them, a tap-to-jump game I called Celo Dash:

Celo Dash, one of my Claude Design playground concepts. Never shipped, just me seeing how far an idea could go.

It does way more than UI

Here’s what surprised me most: Claude Design isn’t just for web and mobile UI/UX. I’ve leaned on it for thumbnails, banners, even logos, and the results are genuinely good.

Logos are the one spot that’s still a bit hit-or-miss. That’s the area I’m hoping gets better over time. But everything else? Better than I’d ship on my own, and faster than I’d get there alone.

A partner, not a magic button

I want to be clear though: it’s not magic, and I don’t treat it like it is. The first version is rarely the final one. Sometimes it plays too safe and bland, sometimes it overdoes an effect. The taste, the “no, that’s not it” instinct, that part is still on me.

But that’s exactly why “partner” is the right word. I bring the judgment and the direction; Claude brings the speed and the thousand tiny decisions I’d otherwise grind through one at a time. We meet in the middle, and we get there a lot faster than I would alone.

Wrapping up

The design stage used to be the slow part of any project. Now it’s almost my favorite, because I can actually play, trying five directions in the time it used to take me to settle on one.

If design is the part that always drags for you, give this a real shot. Stop thinking of it as “generate a UI for me” and start treating it like a conversation with a fast, tireless design partner. That reframe changed everything for me.

P.S. Fittingly, this post was written with a little help from AI too.