The interesting thing about vibe coding isn't what people are building. It's what they're building on.
We spend a lot of time in this business chasing the perfect name. The exact match. The clean .com. The short brandable that passes the radio test, the billboard test, and three other tests somebody invented on a forum twenty years ago.
Vibe coders aren't playing that game.
They're not asking what the best possible domain is for the next ten years. A lot of them are asking one question: what can I launch on today?
And that one question changes everything about the domain decision.
Vibe coding collapsed the distance between idea and live product down to days. Sometimes hours. That doesn't mean the products are good, or secure, or businesses, plenty are none of those things. But the gap between "I had a thought" and "it's online" is basically gone now, and when launch speed is the whole point, the name stops being a ten-year decision and becomes a Tuesday decision.
Look at Pieter Levels and fly.pieter.com. No domainer would draw that up. It's a subdomain on the guy's personal site. And it reportedly cleared $1 million in ARR in 17 days. The domain didn't carry the product, Pieter's audience did. Distribution beat the domain, and it wasn't close.
Then there's Stanley, running on getstanley.ai. You know this pattern by now. Can't get the brand, slap "get" in front of it, ship. We used to call that a compromise. Founders call it Tuesday. Stanley reportedly hit $50K MRR in six weeks, and the .ai is doing real work . It tells you what the thing is before you've read a word of copy.
trypayout.app is the same story with a different prefix. "Try" isn't pretty. But .app fits, and the site now claims 400,000-plus downloads, 300,000-plus claims filed, and $80 million-plus found for users. Did the domain cause that? No. Did it get in the way? Also no.
Then you get the infrastructure layer showing up on names like sheetninja.io, a tool that turns a Google Sheet into an API so somebody on Lovable or Bolt or Replit can ship without building a backend. .io still belongs in that world. And I'll admit I've got a soft spot for the ones where the extension actually pulls its weight, like VibeTalent on vibetalent.work. Would the .com be better? Sure. Is .work killing that project? Come on.
Here's what domainers keep missing about all this.
🔥 Check out NotRenewing.com
If you're dropping names, you might as well get paid for them...
Founders aren't picking names by our scorecard. They're picking on speed, price, availability, and whether the name explains the product in about a second. Good enough to get in front of users beats perfect-but-later every time, because if the thing dies in two weeks, and most do, they didn't bury real money in a premium name for a coin flip.
None of this is brand new. Startups have leaned on prefixes and oddball extensions and just-okay domains forever. What's new is the volume. More experiments, more launches, more naming decisions getting made in public every single day. Which, if you're paying attention, is a firehose of free market data about what builders will actually accept.
And the data isn't subtle. Builders want names that explain the product fast, that are available right now, on .ai or .app or .io or .sh when the extension fits and they'll live without the perfect .com on day one.
What they won't live without is trust.
That's the part I don't want anyone to misread. It's tempting to look at this pile of scrappy launches and decide domains matter less now. I think it's the opposite. Bad names get exposed faster now. Somebody ships a tool over a weekend on a confusing, sketchy-looking name and the market scrolls right past it just as fast as it got built. The domain is still the first impression. It still has to earn the click.
So where does that leave us as investors? I think the boring names get more interesting. Not just the one-word .com unicorns. The names that match a specific small tool, a calculator, a directory, a narrow SaaS workflow. Vibe coders are building ten thousand specific things, and specific names serve specific things. Alternate extensions deserve an honest look when they add meaning too, though most random .ai is still junk and you already know it. And that "get/try/use/join" version of a name a founder launches on? That founder might be your upgrade buyer in eighteen months, once there's traction worth protecting. The first domain is a test. The better one gets attractive the day the test passes.
I'm not telling you to go register every AI-flavored phrase that sounds like a weekend project. Most of these things will never make a dollar. A lot will be half-built, insecure, or forgotten by Friday.
But the clues about where naming is headed aren't coming from our forum arguments. They're coming from what builders are doing while we argue. Right now they're telling us speed and clarity get you in the door, and the perfect domain can wait.
And then the second something actually works, watch how fast that domain starts to matter again.