If you place a 3D model of a car using the No-Code Editor and then write code in the Code Editor, that code cannot control or modify the car. You can't make the car's wheels spin, change its color, or move it around using your custom code. The car exists in the No-Code Editor's space, and your code exists in the Code Editor's space.
If you want to add custom programming to any content, whether that's a 3D model, video, image, or anything else, you need to load that content inside the Code Editor using three.js, which is a popular tool for creating 3D graphics on the web.
Imagine you have a 3D model of a product, perhaps a sneaker for a shoe company's AR experience. You want this sneaker to respond to user interactions in a unique way. When someone taps different parts of the shoe, you want information panels to appear with details about those specific features. The No-Code Editor's standard tap behaviors might not support this level of detail.
To make this work, you would load the sneaker model directly in the Code Editor rather than placing it through the No-Code Editor. Inside the Code Editor, you use three.js to import the model, and then you write code that detects where the user taps and responds accordingly. All of this logic lives together in the Code Editor, which means it all gets packaged into that single file when you save to the No-Code Editor.
The key insight is that custom code can only control content that lives in the same space. By loading your sneaker in the Code Editor, you bring it into the same space as your code, making full customization possible.
This applies to any type of content. Videos that need custom playback controls, images that need to transform in specific ways, 3D models that need complex animations, all of these should be loaded in the Code Editor if you want to control them with custom code.