Niantic’s transformation is one of the more interesting pivots in the XR space right now. In late 2025 the company sold its gaming division — home to Pokémon GO, Pikmin Bloom, and the rest of the location game portfolio — to Scopely for around $3.5 billion. What it kept was the underlying technology: a decade’s worth of location data, real-world mapping infrastructure, and spatial computing platform tools. Then it spun that out as Niantic Spatial with $250 million in fresh funding.
The question for developers is whether this is just a restructuring story or whether Niantic Spatial represents something genuinely worth paying attention to for location AR work.
Here’s the thing: the asset Niantic actually built over ten years isn’t the games. It’s the map.
What Niantic actually has
Niantic’s location game portfolio required players to physically go to places. That incidentally generated one of the most comprehensive real-world mapping datasets in existence — millions of verified real-world points of interest, semantic location metadata, pedestrian pathways, and crowd-sourced anchor data at a granularity that neither Google Maps nor Apple Maps was built to provide.
That data underlies Niantic’s Visual Positioning System (VPS), which lets AR experiences anchor precisely to real-world locations — not just GPS coordinates, but specific spots at specific orientations. A user pointing their phone at a pub sign gets an AR overlay that’s anchored to that exact sign, not a floating label positioned by GPS alone.
The Lightship platform, Niantic’s developer SDK, exposes this infrastructure. AR apps built on Lightship can use VPS for precise real-world anchoring, mesh generation (building 3D representations of the physical environment from device camera data), occlusion handling, and multiplayer spatial experiences where multiple users see shared AR objects in the same physical space.
The Niantic Spatial pitch
Post-spin-off, Niantic Spatial’s pitch to developers is cleaner than it was when the same company was also running Pokémon GO. It’s now explicitly a B2B infrastructure play: give us access to your app’s camera data and real-world movement patterns, and in return you get access to the best real-world AR anchoring infrastructure available.
The use cases it’s pitching to enterprise customers include: location-based brand activations, cultural heritage experiences (museums, tourist attractions), retail AR in physical stores, and urban wayfinding that needs more precision than GPS allows.
For enterprise AR specifically, the VPS advantage is meaningful. GPS gets you to the right building. VPS gets you to the right aisle, the right exhibit, the right product shelf.
The developer experience in 2026
Lightship ARDK (Augmented Reality Development Kit) is the primary integration point. Unity SDK is mature and well-documented. Unreal support exists but has historically been less maintained. The web SDK for WebXR experiences is more limited — browser-based location AR remains awkward everywhere, not just on Niantic’s platform.
The VPS coverage map is the most practical consideration. VPS anchoring works where there’s scan data — urban centres, popular landmarks, and areas where Niantic games drove data collection. Outside these zones, you fall back to standard GPS and device-based SLAM. Coverage is dense in major UK cities, tourist destinations, and retail hubs, sparse in rural and suburban areas.
The SDK licensing is generous for smaller developers and tightens at scale — which is a reasonable model for infrastructure that has real ongoing costs.
What it means for the location AR ecosystem
The separation of the games business from the platform business matters more than it might look. When Niantic was primarily a games company, the platform was a side project and enterprise deals were secondary to the next game launch. As a dedicated infrastructure company, the Lightship roadmap isn’t competing for resources with Pokémon GO updates.
Whether that translates to faster developer tooling, better API stability, and more reliable enterprise SLAs remains to be seen over the next 12 months. The $250M funding gives Niantic Spatial runway to operate independently through 2028 at typical burn rates, which is enough to either establish the infrastructure business or demonstrate it isn’t viable.
For developers working on location AR: Niantic Spatial’s VPS is still the most credible real-world anchoring solution available outside of building your own mapping infrastructure. The iOS and Android ARKit/ARCore pipelines have improved, but they don’t have the pre-built location data layer. If your use case depends on sub-metre AR anchoring at specific real-world locations, Lightship remains worth the evaluation.
For those waiting on the broader ecosystem: the Niantic Spatial spin-off is a vote of confidence that location AR as infrastructure has a standalone business case. That’s a signal the space is maturing past the games-and-demos phase.