Innoactive Spatial includes a built-in USDZ viewer for opening and exploring 3D models directly on the device. It is a self-contained alternative to the workstation-streamed workflow: you load a .usdz file (from local storage or a web link) and the model renders natively on the headset or phone, no Innoactive Spatial Runtime or RTX workstation required.
Innoactive Spatial USDZ viewer vs Apple Quick Look
The viewer is a custom-built application rather than the system-provided AR Quick Look. Because it renders with the same underlying engine as Quick Look — Apple's RealityKit — models look identical between the two. What the Innoactive Spatial viewer adds on top of Quick Look:
- Shared multi-user viewing. A host can share a model and every participant sees the same thing in the same place.
- Persistent placement. Anchor a model in a room, on a surface, or to an image marker so it stays where you put it across sessions.
- Custom manipulation gestures. Our own scaling, rotation, repositioning, and height-adjustment gestures rather than the system defaults.
- Logitech Muse pen support. Use the pen as a laser pointer, draw annotations on or around the model, and erase them — none of which Quick Look offers. See Point and annotate with the Logitech Muse pen for setup details.
- Asset caching. Models you have opened before re-open quickly, with downloads handled transparently.
How it works under the hood
- Engine: the viewer is built on RealityKit, Apple's first-party rendering engine, on both Apple Vision Pro and iPhone or iPad. Models you have validated for Quick Look will render the same way in the Innoactive Spatial viewer.
- MaterialX shaders: on Apple Vision Pro, MaterialX shaders embedded in a USDZ render automatically. On iPhone or iPad this is an Apple platform limitation — the model falls back to its standard PBR materials, so MaterialX-specific effects will not appear.
- No streaming dependency: loading and rendering happens on the device itself. Different from the streamed workflows for VRED, Unreal Engine, Unity, or Omniverse, no workstation is involved.
What the viewer can do
- Loads and displays USDZ 3D models directly on the device — no streaming or workstation required.
- Works on both Apple Vision Pro (fully immersive) and iPhone or iPad (windowed AR).
- Opens models from your own files, from web links, and includes built-in sample models.
- Multi-user viewing — the host shares a model and everyone sees the same thing.
- Persistent placement — anchor a model in a room, on a surface, or to an image marker.
- Natural manipulation gestures — rotate, pinch-to-scale, reposition, and adjust height.
- Logitech Muse pen support — use the pen as a laser pointer, draw annotations on or around the model, and erase what you have drawn.
- MaterialX / advanced shader rendering on Apple Vision Pro.
- Smart caching and download so repeat viewing is fast.
- Memory safeguards that warn before opening very large files.
Limitations
- File size depends on the device. There is no hard cap. How large a model the viewer can open and render smoothly depends on the Apple Vision Pro generation (M2 vs M5), the triangle count, and the number and complexity of materials in the file. The viewer includes memory safeguards that warn before opening very large files — beyond that, validate with your own models on the headset(s) you intend to use.
- No interactions yet. Models render statically. Element-level interactions — opening car doors, toggling lights, triggering animations from user input — are not supported by the viewer at the moment.
Related articles
- Point and annotate with the Logitech Muse pen — pen setup, gestures, and annotation workflows.
- Anchoring methods — pick the right anchoring method (world, image, object, Logitech Muse pen) when placing a USDZ model in a room.
- Spatial collaboration with co-located and remote teammates — how multi-user sharing works across participants.