AR gaming has been "almost ready" for eight years. ARKit 7 is the version where the platform finally caught up.
Pokémon GO proved AR gaming could be a hit and then nothing else really did for a long time. The platform was not quite there. ARKit 7 (and ARCore's parallel evolution) closed the gaps — persistent objects, shared sessions, and cm-scale placement work reliably for the first time.
Persistent AR
Objects placed in the world stay there across sessions. Apple's Object Anchors and Google's Cloud Anchors both store enough geometric information to relocalize at the same physical spot. The architecture for "place a virtual basketball hoop in your park" finally works.
Shared multiplayer AR
Two devices in the same space can see the same virtual content with sub-centimeter alignment. Niantic's Lightship and Apple's Multipeer ARKit both deliver this. The use cases are concrete: tabletop strategy games, location-based PVP, AR escape rooms.
Hand and body tracking
Vision Pro's hand-tracking, paired with iPhone Pro models that have LiDAR + a TrueDepth front camera, makes "use your hands as the controller" reliable. Quest 3 does the same on the headset side. Free-form gestural input is no longer a research demo.
Scene understanding
The latest ARKit and ARCore APIs return a semantic scene graph: this is a wall, this is a floor, this is a table, this is a chair. Game logic can now condition on the player's environment. A monster spawns from the closet, not from the wall.
What is now feasible
- Location-based MMOs with persistent shared world state.
- Tabletop strategy games where the table is your kitchen table.
- Fitness games that adapt to the player's actual room layout.
- Escape-room-style narrative games tied to specific real-world locations.
What is still hard
Battery and thermal limits — sustained AR sessions still cook devices. Outdoor occlusion (do trees block your spell?) is unreliable. Indoor lighting still defeats some scene understanding. And the user experience problem of "people don't want to hold a phone in front of their face for 30 minutes" has not gone away.
The path forward
The next inflection is glasses, not phones. Vision Pro is a preview; the consumer version of glasses-form AR is the breakout opportunity. Studios building AR games today should be building for that future device, not the current one.