I created the world's first 360° spatial AI platform on Apple Vision Pro - built natively for visionOS, never owned the device [P]

https://preview.redd.it/sy5m7v20smwg1.png?width=2914&format=png&auto=webp&s=4b327a3c5e9db30d5b556a6575c4761f53b9cb1e

https://preview.redd.it/a0m08v20smwg1.png?width=2810&format=png&auto=webp&s=2e8dad47e8061a1c98f3b5c6132c3020ffae1f55

I want to show you something nobody has ever seen before.

Three months ago I had zero coding knowledge. I couldn't write a single line of code. In the time since, I taught myself GitHub, Visual Studio, Xcode, Android Studio, Firebase, Firestore, Vercel, Sentry - and built a fully functional AI platform live across web, iOS, Android, Mac desktop, and Apple Vision Pro.

I have spent approximately 3 months spending 16 hours a day working on this project to get it to where it was on web, android and iOS.

Today I converted it into something completely new.

Asksary is a world-first fully spatial AI experience - built natively for visionOS. Not an iPad app running in compatibility mode. A ground-up, native spatial build where the entire interface is a live immersive 360° wallpaper. You don't open the app. You step inside it with realtime voice chat with OpenAI WebRTC with 8 voices with near zero latency too.

In the first screenshot you'll see GPT-5 greeting you from inside the spatial environment - all happening inside a 360° world with floating UI, particle effects, and a starfield you're literally standing in.

The screenshot shows how Realtime voice chat looks like. Put on the Vision Pro. Change the 360 spatial experience background and chat with OpenAI with near zero latency in realtime.

It currently runs GPT-5.2, Claude Sonnet 4.6, Grok 4, Gemini Ultra, DeepSeek R1, 01 Pro

30 live interactive wallpapers and themes. Each one is a different world to inhabit while you work.

Beyond the spatial shell, the platform includes:

• Image generation via GPT-Image-1 and Nano Banana Pro
• Flux Image Editor with visual history
• Video Studio - Luma Dream, Veo 3.1, Kling 1.6, 2.6 and 3, up to 10 second AI videos with audio - view in full screen on the Vision Pro Display.
• Music Studio - 30 second tracks via ElevenLabs with custom visualiser
• 3D Model Studio with STL export (coming soon) - 3d model something in an ever immersive space with full 360 view.
• Vision to Code - screenshot any UI, get live editable code that can be viewed in spatial space. Move around, resize etc
• Web Architect, Game Engine, Code Lab - all on a live canvas with instant run features
• Real-time 2-way voice chat, Podcast Mode, Voiceover with openAI WebRTC
• Full productivity suite, business tools, social tools, UI interface with full RTL support and 26 languages
• 18 API integrations total
• Persistent cross-model memory, start on Grok on your phone. Pick up the Vision Pro and continue in Claude without having to re-explain anything. It just knows your previous message history no matter what device or platform your using.

I wanted to build something that made people say wow. Something nobody had done. I think this might be it.

I did this without ever having a Vision Pro at hand to help me develop the concept. So I've never experienced it for myself but I have a pretty good imagination to what it would be like.

This version of the Apple Vision Pro variant is not currently available on the App Store but if people are genuinely interested I'll release it soon enough.

Would love to hear what you think of the whole idea. It's a fully working model, so not a prototype or demo either.