In short: AR overlays, MR reacts, spatial computing is the umbrella term above both, and VR sits inside that umbrella as the fully synthetic case with no real-world blend at all.
The quick answer
These three terms get used interchangeably in briefs and pitches, and that is where confusion, and the wrong quote, comes from. Here is the fast version before the detail:
| Term | What it does | Example |
|---|---|---|
| AR | Overlays digital content on the real world | A Snapchat lens, a phone measuring app |
| MR | Understands the real world and reacts to it | Content that hides behind a real object |
| Spatial computing | Umbrella term for any 3D-space-aware system | Covers AR, MR, and VR together |
Augmented reality (AR): overlay only
AR places digital content on top of a view of the real world, usually through a phone camera or a pair of glasses. The key word is overlay: the digital content sits on the real world without necessarily understanding it in depth. A face filter tracking your features, a Snap Lens placing a graphic on a table, a phone app that draws an arrow toward a destination, all of these are AR. It is the lightest-weight case of the three, and often the fastest and cheapest to build.
Mixed reality (MR): overlay that understands and reacts to the room
MR does everything AR does, plus real-time understanding of the physical space well enough that digital content behaves like it belongs there. That means occlusion, a virtual object correctly hiding behind a real one, persistence, content staying anchored to a spot across a session, and physics-like interaction, a digital ball that appears to bounce off a real wall. Apple Vision Pro and the spatial computing SDK behind Snap Spectacles both operate at this level.
The practical difference between AR and MR in a brief is usually the difference between "put a graphic on the screen" and "make the content aware of the room it's in." The second is a meaningfully bigger build, and briefing it as the first produces underscoped quotes.
Spatial computing: the umbrella term
Spatial computing sits above both. It is any system that computes with an understanding of 3D space, rather than a flat screen, whether or not it blends that space with reality at all. AR and MR both fall inside it, and so does VR, which replaces the physical world entirely with a synthetic one but still tracks the user's position and movement within that 3D space. A fuller explainer on the category itself is at what is spatial computing.
Spatial computing is not a fourth type of reality alongside AR, MR, and VR. It is the category name for all three of them put together.
Where Snap Spectacles, Meta Ray-Ban Display, and Apple Vision Pro each sit
- Snap Spectacles, marketed as AR, but its spatial anchors and depth understanding give it genuine MR capability, see our own Spatial Mapper and Noodle builds for what that looks like in practice
- Meta Ray-Ban Display, closer to pure AR: a display overlay and AI assistant, without the deep spatial mapping Spectacles' SDK provides
- Apple Vision Pro, spans the whole range, full VR in fully immersive apps, MR when passthrough blends real and digital content, AR-like when windows pin to real surfaces
A full spec-level comparison of the smart glasses side of this sits in our smart glasses comparison.
Why the distinction actually matters for a brief
The wrong term in a brief tends to produce the wrong quote. Asking for "AR" when the actual need is content that reacts to a specific physical room, occlusion behind real furniture, a setup that persists across a whole event day, describes MR-level work, which costs and builds differently from a simple camera overlay. Naming the capability you actually need, not just the buzzword that sounds closest, is the fastest way to get an accurate quote the first time. Our own approach to spatial and wearable briefs sits at our wearables and smart glasses page, and the live demos at ar.rbkavin.studio let you feel the AR-to-MR spectrum firsthand before writing a brief at all.
Frequently asked questions
What is the difference between AR, MR, and spatial computing?
AR (augmented reality) overlays digital content onto a view of the real world. MR (mixed reality) goes further, understanding the real world well enough that digital content can react to it, sitting behind a real object or bouncing off a real wall. Spatial computing is the umbrella term above both, covering any system, AR, MR, or VR, that computes with an understanding of 3D space rather than a flat screen.
Is Snap Spectacles AR or MR?
Snap Spectacles is generally described as AR hardware, but its spatial computing SDK gives it real MR capability: spatial anchors, hand tracking, and depth understanding let content react to the real room, not just sit overlaid on top of it. The Spectacles case studies at RBKAVIN, Noodle and Spatial Mapper among them, both use that MR-level understanding of physical space.
Is Apple Vision Pro AR, MR, or VR?
Apple describes Vision Pro as a spatial computer rather than committing to one label, and it genuinely spans the range: full VR when the display shows only digital content, MR when passthrough video blends real and digital content that reacts to the room, and AR-like behavior when app windows are pinned to real surfaces. It is a strong example of why spatial computing works better as the umbrella term than any single label underneath it.
Why does the difference between these terms matter for a brief?
Because the wrong term in a brief tends to produce the wrong quote. Asking for "AR" when the actual need is content that reacts to a specific physical room, occlusion behind real objects, persistence across a session, describes MR-level work, which costs and builds differently than a simple camera overlay. Naming the right capability, not just the right buzzword, gets a brief answered correctly the first time.
Where does VR fit into spatial computing?
VR (virtual reality) is fully synthetic: the real world is replaced entirely by a digital environment, with no camera passthrough or overlay. It sits inside spatial computing as the umbrella term because it still computes with 3D space and tracks the user's position and movement within it, it simply does not blend that space with anything physical.
Insights newsletter
Smart glasses, AR campaigns, spatial computing.
Straight to your inbox. No noise.
SubscribeNot sure which term describes your brief?
Tell us what you want the experience to actually do. We'll tell you whether that's AR, MR, or something else, and what it costs to build.
Start a project