
Open-source Meta Quest 3 app that applies real-time AI style transformation to passthrough camera footage at 150–200ms latency.
Some links may be affiliate links. We may earn a small commission at no extra cost to you. Learn more

Open-source Meta Quest 3 app that applies real-time AI style transformation to passthrough camera footage at 150–200ms latency.
Category
AI Simulation
Decart XR is free to use. The client app is open-source under MIT license on GitHub and available as a precompiled sideload via SideQuest. The Decart AI service that processes the video transformation is governed by separate terms at decart.ai. Requires a Meta Quest 3 headset.
| Plan | Details |
|---|---|
| Free | Free to use. App available via SideQuest for Meta Quest 3 sideloading. Client source code on GitHub under MIT license. Decart AI service terms apply separately. |
Quick Summary
Decart XR is an open-source AI application for Meta Quest 3 that applies real-time video-to-video style transformation to the headset's passthrough camera feed, letting users see their physical surroundings reskinned in one of 61 premade visual styles — from Studio Ghibli and cyberpunk to fantasy and medieval environments — or described through custom voice prompts. Developed by Decart AI and released in October 2025, the app processes live camera frames at 30fps and 720p resolution with approximately 150 to 200 milliseconds of end-to-end latency. The source code is publicly available on GitHub under MIT license, and the precompiled app can be sideloaded on Meta Quest 3 via SideQuest.
Associated Tags
Meta Quest 3 AI, VR style transformation, real-time AI video, passthrough AI app, open-source VR AI, video to video AI, mixed reality AI
Discover practical workflows and real-world scenarios where Decart XR delivers key solutions.
Experiencing a familiar room reskinned as a cyberpunk environment, anime world, or medieval scene by pointing a Quest 3 at the physical space and selecting a preset style
Providing a custom voice prompt describing a specific visual environment and watching the passthrough view transform toward that description in real time
Using the application as a technical reference for building Meta Quest 3 apps that combine passthrough camera access, WebRTC streaming, and AI video processing
Demonstrating real-time AI video-to-video transformation capabilities in a VR context at developer meetups, research showcases, or technology events
Forking the open-source client on GitHub to experiment with modified WebRTC parameters, custom style models, or extended camera access implementations
Testing the practical latency and consistency of AI style transformation on consumer VR hardware as part of an XR research or feasibility study
Reviewed by Sohail Akhtar
Lead Editor & Founder
What we like
Limitations
Who should use Decart XR?
Open-source AI world model by Decart and Etched that generates real-time Minecraft-style interactive gameplay at 20 FPS using next-frame prediction, with no traditional game engine required.
World Labs' real-time generative world model that creates persistent, navigable 3D scenes from a single image on one H100 GPU.
Tencent's open-source AI model that generates interactive, action-controllable game video sequences from a single image and keyboard inputs.
Multimodal AI world model by World Labs that generates persistent, navigable 3D environments from text, images, video, or 3D layouts, with in-scene editing and Gaussian splat, mesh, and video export.