Niantic Spatial
"Real-world foundation models for physical AI"
Builds real-world foundation models and a Visual Positioning System that give robots and AI centimeter-accurate localization and 3D understanding, including.
Niantic Spatial, Inc. builds real-world foundation models for physical AI, enabling robots, agents and autonomous systems to operate using world models grounded in physics and geometry. The company spun out of Niantic in May 2025, retaining the geospatial AI division and the Scaniverse and Lightship technologies, and is headquartered in San Francisco.
Its platform is trained on more than 30 billion posed images plus terrain, drone and satellite data and powers three core capabilities. Reconstruct builds geometrically accurate digital twins from smartphones, 360-degree cameras, drones, satellites and LiDAR, output as meshes and Gaussian splats; the company created SPZ, the open-source file format for Gaussian splats now used across the industry. Localize is a Visual Positioning System that turns ground and aerial capture devices into a precise positioning and orientation sensor, delivering 6DoF localization with near-centimeter accuracy almost anywhere, including GPS-denied environments, powered by 50 million trained neural networks and 150 trillion parameters. Understand provides spatial intelligence, an AI interface that can query, measure and reason about every point in 3D. The company serves robotics (continuous localization for robots indoors and in GPS-degraded settings), defense and intelligence, and oil and gas.
It has filed or holds 140+ patents and publishes at top computer-vision conferences such as ICCV and ECCV. Customers and partners include Coco Robotics, Spexi and Snap. Founder John Hanke serves as executive chairman, with Inhi Cho Suh as CEO since March 2026.
Primary type & automation activities this supplier delivers:
Localize (Visual Positioning System)
Vision-based 6DoF localization with near-centimeter accuracy, including in GPS-denied environments.
Reconstruct
Creates geometrically accurate digital twins from phones, 360 cameras, drones, satellites and LiDAR.
Understand
Spatial intelligence AI interface that can query, measure and reason about every point in 3D.
Scaniverse
3D capture app and front door to geospatial AI services for high-quality scanning.
Contact Niantic Spatial
WEBSITE
https://www.nianticspatial.com/HEADQUARTERS
United States
Company Facts
Founded
2025
Primary Role
Software/Algorithm
Company Size
-
Primary Region
North America
Annual Sales
-
Funding Stage
-
Funding Total
-
Certifications & Memberships
Niantic Spatial adds USDZ export to Scaniverse to streamline robotics simulation workflows
Niantic Spatial has launched USDZ export for its Scaniverse app, enabling robotics developers to convert real-world environments into simulation-ready digital twins for use with Nvidia Isaac Sim. The new capability is designed to help address the long-standing “sim-to-real” gap in robotics, where systems trained in synthetic environments often struggle when deployed in complex, real-world settings. […]
News Robot simulation 3d scanning Autonomous robots digital reconstruction digital twins
Pokémon Go creator wants your drone photos to train AI
Niantic Spatial, the spinoff company from the creators of Pokémon Go, is investing heavily in drones and artificial intelligence by partnering with Spexi, the largest decentralized drone imagery network. This collaboration aims to develop intricate 3D maps of the real world, which are intended to enhance the training of future AI systems and robots. The initiative reflects Niantic's commitment to leveraging advanced technologies to innovate and expand its capabilities beyond mobile gaming.
News
Niantic Spatial and Spexi Turn Drone Imagery Into Physical AI
Niantic Spatial and Spexi Geospatial have formed a strategic partnership to enhance the use of aerial drone imagery for enterprise applications. This collaboration integrates Spexi’s extensive aerial data network with Niantic Spatial’s advanced 3D reconstruction technology, enabling the creation of city-scale 3D Gaussian splats. Announced recently, this initiative aims to provide businesses with on-demand access to drone imagery, facilitating the development of physical AI solutions. The partnership is expected to streamline the process of transforming aerial captures into actionable data for various industries, marking a significant advancement in the utilization of drone technology for commercial purposes.
AI Drone News Drone News Feeds Mapping News 3D reconstruction
