New iPhone App Performs 3D Reconstruction for AI
The SpAItial Echo-3 app allows users to create 3D digital twins using standard video, replacing heavy scanning hardware.
Updated on Oct. 28, 2026 in Artificial Intelligence
Live Poll
Do you find more personal value in digitizing real-world spaces or creating entirely new virtual environments?
SpAItial has released Echo-3, an iOS application that generates 3D digital twins from iPhone video recordings. The software is available now for devices running iOS 18 or later and produces assets in formats including meshes and point clouds.
Why it matters
By enabling 3D reconstruction from consumer hardware, SpAItial aims to lower the barrier for gathering real-world data required by physical AI systems. This move challenges the traditional reliance on expensive professional-grade laser scanning equipment.
The application supports the generation of 3D Gaussian splats, point clouds, and meshes, maintaining real-world metric scales in the exported assets. Scans require five to ten minutes of footage to process.
The players
SpAItial
A startup focused on spatial AI and 3D reconstruction technology that raised $13 million in seed funding.
Matthias Niessner
The CEO and a co-founder of SpAItial who oversees the company's research and product development strategies.
The details
Echo-3 utilizes Gaussian splatting, a method that represents 3D scenes as a collection of thousands of small, overlapping 3D Gaussian kernels. These kernels store color and opacity data to enable high-fidelity visual reconstruction from 2D images. The app automates the alignment and geometry estimation process, which previously required manual stitching or expensive lidar-based hardware setups.
Timeline
May 2025: SpAItial secured a $13 million seed funding round.
Spring 2026: The company released its previous software iteration, Echo-2.
October 27, 2026: SpAItial launched the Echo-3 application for iOS.
The Tech Race
SpAItial is positioning its software to compete with existing photogrammetry and lidar-based scanning workflows. This launch marks an evolution toward Gaussian splatting for real-time spatial data capture, moving beyond the academic research that defined early 3D reconstruction benchmarks.
Users with an iPhone running iOS 18 can now create 3D digital twins by recording five to ten minutes of video. Access is gated by a subscription plan requiring credits, though specific pricing for those credits is currently unannounced.
The takeaway
SpAItial is moving 3D reconstruction from specialized labs to consumer mobile devices to accelerate the training of physical AI. Watch for future performance benchmarks comparing Echo-3 outputs against professional laser-scanning results.
What happens next
The company has announced that Android support is planned for a future release.
Further reading
For broader trends in 3D spatial modeling, visit the Artificial Intelligence section.
Source note: This article includes information reported by Startup Fortune.
Live Poll
Do you find more personal value in digitizing real-world spaces or creating entirely new virtual environments?
