appless
What if your phone had no apps
What is it?
What it is
AppLess is a mobile application that demonstrates a generative UI operating system. It generates screens on demand from natural language prompts, streams the UI description (openui‑lang) from a large language model, and renders native components (Cupertino on iOS, Material 3 on Android) in real time.
Why it exists
It exists as an experimental showcase of what a generative‑UI OS could feel like, proving that a single component contract can drive a model to produce fully interactive native screens without any pre‑built code or backend services.
Who should use it
Developers and designers exploring generative UI concepts, React Native enthusiasts, AI researchers interested in LLM-powered interfaces, and open‑source contributors wanting to prototype a generative‑UI operating system.
Who should avoid it
Users looking for a production‑ready, fully integrated mobile app solution, non‑technical stakeholders, or those who need immediate, real‑world data integration without custom wiring.
How it works
A quick walkthrough in plain English
How appless works
Step 1 of 3
You interact with it
Open appless, send a request, or connect it to your stack.
Features
Advantages
- Eliminates traditional app menus and home screen icons
- Native look and feel without prompt modification
- No backend deployment required (direct-to-model)
- Incremental rendering of UI as tokens stream
- Extensible architecture for real-world service integrations
Disadvantages
- Experimental nature with no real service integrations
- Simulated actions (e.g., payments and bookings do nothing)
- Requires user-provided API keys (e.g., Cerebras)
- Content is plausible fiction unless web tools are enabled
Installation
native
git clone https://github.com/thesysdev/appless.git cd appless npm install npm run ios || npm run android || npm start
FAQ
What is AppLess?
AppLess is an experimental generative UI operating system demo. Instead of traditional apps, it uses an LLM to generate and stream interactive native UI components (Cupertino on iOS and Material 3 on Android) on demand based on user requests.
Does the app require a backend to run?
No. Generation runs directly from the phone to the model. On first launch, the app asks for a Cerebras API key, which is stored securely on-device using Keychain or Keystore.
Are the actions and data in the app real?
No. AppLess is an experiment. While it can use tools like Exa for web search or Unsplash for images to ground content in real data, all functional actions (like 'Order placed' or 'Flight booked') are simulated and do not affect real-world services.
How does the UI generation work technically?
The app uses the OpenUI stack. An LLM writes 'openui-lang' (a streaming-first DSL), which is then parsed by the @openuidev/react-lang runtime and rendered into real native components via a component contract defined in TypeScript.
How can I enable real-time web search or semantic photos?
You can enable these features by adding the corresponding API keys to a `.env.local` file in the project root: `EXPO_PUBLIC_EXA_API_KEY` for web search and `EXPO_PUBLIC_UNSPLASH_ACCESS_KEY` for semantic photos.
Featured in Videos
YouTube tutorials and walkthroughs for appless
Alternatives
Similar projects ranked by category, topics, and text overlap.
