Ada Is Live

Ada is officially available to everyone. Describe your app in a few words and Ada builds the real thing — the screens, sign-in, and a database designed around your app — landed directly on the canvas, ready to direct. We launched it alongside a completely new adalo.com, because the product changed enough that the old story no longer fit.

Type it. Build it. Ship it.

This one is personal. Ada started as my idea. Before there was a roadmap item or a name, there were prototypes and blueprints — rough sketches of how an AI could build onto our canvas instead of beside it. Watching that turn into something every builder can open today is the best part of this job.

An app is only the beginning

Since 2019, makers have built more than 3 million apps with Adalo. Watching them taught me something AI hype tends to skip: generating code has never been easier, but building a business has never been just about the code.

What matters is what people do with the app. A landscaping business that runs on it. A garage. A startup that began as a spreadsheet. So Ada doesn't stop at generation:

• Payments connect in a few steps.
• Publishing Studio gets apps store-ready.
• Code export means what you make stays yours — more on why that matters in Why AI App Builders Should Give You Real Code.
• Real iPhone and Android apps, plus web apps that work on desktop.

We're not just making it easier to create software. We want to make it easier for anyone to turn an idea into a real business.

The idea was never "AI in a text box"

The obvious version of an AI builder is a chat window that spits out an app. I didn't want that. I'd watched too many people argue with a prompt box over something they could have fixed in one click if they could just touch it.

So the blueprint for Ada had one rule from the first prototype: the AI does the cold start, and then it hands you the controls. Ada generates the app onto Adalo's visual canvas, and you correct it by pointing at what's on screen — not by re-describing it. I wrote more about that bet in Point, Don't Prompt.

Generation is the easy half. The product is how cheaply you can correct what the AI made.

Lighter, not just smarter

Makers asked us for two things: an easier way to build, and apps that scale and run faster. Ada ships on top of a full infrastructure rehaul, and we held its output to a performance bar from day one:

• 20% less CPU — new apps generated with Ada use 20% less CPU than apps from our previous builder.
• Up to 32% less CPU while typing — the moments where builders and their users feel lag the most got the biggest gains.

Those numbers matter more to me than any launch-day screenshot. Efficiency is the difference between an app people try and an app people keep.

What's next

General availability isn't the finish line — it's the point where real builders start teaching us what Ada should do next. We'll keep pushing the same principles that shaped the first prototype: real apps, a canvas you can touch, and AI you direct rather than obey.

Describe the app. Ada builds it. You ship it.