You will never install an app in the future. Want proof? Here is a demo (not sped up, I promise) built with OpenUI.com and Cerebras. No App Store. No downloads. No predefined screens. Every screen on your phone generated live, the moment you ask.

We’re calling it AppLess.

Here’s what makes it possible:
▪️OpenUI streams directly from Gemma 4 31B running on Cerebras.
▪️Each screen is roughly 300 tokens.
▪️Cerebras does around 1800/second. Thats sub-second rendering.
▪️Even with tool-call turns, that is enough to render every screen natively.
▪️Each screen is personalized per user's context in real time.

This is the part that excites me most. Right now, AI is making us faster at building apps. But AppLess asks a more important question:

Do we need an app for EVERYTHING ?

What if software is generated only when you need it?
Just intent in, live interface out.
That’s the future of Generative UI, delivered today at Thesys.