MultiSet AI v1.9.1–v1.9.3: XGrids & NavVis Support, Matterport MatterPak On‑Cloud, and Cross‑Platform VPS Performance

San Francisco, USA — November 24, 2025 — MultiSet AI, the enterprise visual positioning and spatial mapping platform, today announced three rapid releases — v1.9.1 (September 29, 2025), v1.9.2 (October 9, 2025), and v1.9.3 (November 16, 2025) — that broaden scan‑agnostic ingest (NavVis and Xgrids), add on‑cloud Matterport MatterPak processing, introduce headless REST APIs, and elevate cross‑platform performance across Unity and Meta Quest. Together these updates advance MultiSet’s mission to deliver centimeter‑level, 6‑DoF localization with low drift at enterprise scale.

“Enterprises don’t want lock‑in—they want results. With 1.9.x, we leaned into scan agnosticism and headless APIs so teams can turn any capture—NavVis, Xgrids, Matterport—into precise VPS maps that work across Unity, web, mobile, and XR headsets. The outcome is faster iteration, higher reliability, and localization accuracy we’re proud to quantify at 5 cm median in real deployments.” — Nikhil Sawlani, CEO & Co-Founder, MultiSet AI

XGRIDS K1 Visual Positioning System

What’s new

v1.9.1 — September 29, 2025

• Unity Editor: Localization Simulation - Test and iterate entirely inside Play mode using recorded “simulation data” from the updated MultiSet mobile app. Developers can validate localization logic without device builds or field runs.

  • Docs: https://docs.multiset.ai/unity-sdk/localization-simulation

• NavVis ingest (E57 with equirectangular panoramas) - Directly import NavVis outputs (e.g., VLX, M6) to create large‑scale, AR‑ready maps—another step toward fully scan‑agnostic VPS creation.

  • Docs: https://docs.multiset.ai/basics/third-party-scans/navvis-scans

• REST API for E57 uploads - Headless map creation via a multipart upload flow—ideal for platform integrations and white‑label pipelines.

  • API: https://docs.multiset.ai/basics/rest-api-docs/map-upload-api

• MultiSet Mobile app update - Capture Localization Simulation data for the Unity workflow described above.

v1.9.2 — October 9, 2025

• On‑Cloud processing for Matterport MatterPak - Create VPS maps from MatterPak files directly in the MultiSet Developer Portal—no local high‑end hardware required.

  • Guide: https://docs.multiset.ai/basics/third-party-scans/matterport-e57/matterpak-files

• MatterPak REST endpoint (OAuth or client credentials) - Kick off processing programmatically; MultiSet fetches and processes the MatterPak in the cloud.

  • API: https://docs.multiset.ai/basics/rest-api-docs/map-upload-api/matterpak-upload-api

• Meta Quest SDK updatesLocalization simulation in Unity Editor, a new object‑tracking sample scene, and multi‑frame localization for more robust, sticky performance in dynamic environments.

  • Repo: https://github.com/MultiSet-AI/multiset-quest-sdk

v1.9.3 — November 16, 2025

• Xgrids scanner support (E57 export) - Official guidance to convert and ingest Lixel/Xgrids captures via E57 export—further expanding scan compatibility.

  • Docs: https://docs.multiset.ai/basics/third-party-scans/xgrids-scans

• Map analytics: total mapped area (sq ft) - Teams can now quantify spatial coverage and rollout progress at a glance.

• Performance & stability improvements

  • Unity SDK: fixed Mapping Scene upload issue.

  • Mesh generation: improvements for E57 imports.

  • Android 16 (API level) support via updated glTF and Draco packages.

  • Developer Portal: mesh visualization optimizations for large Maps/MapSets.

  • MapSet manual alignment transform fix.

  • MultiSet App UI optimized for iPad.

    
    

Read Full Changelog

Why it matters — Xgrids, NavVis, Matterport VPS

This release strengthens our Xgrids NavVis Matterport VPS workflows across ingest, APIs, and runtime performance.

• Scan‑agnostic VPS: Use the capture tools you already own — NavVis, Xgrids, Matterport — to build reliable AR experiences across indoor and outdoor environments, varied lighting, and at multi‑floor scale.

• Headless & fast: REST endpoints enable automated, white‑label pipelines. On‑cloud MatterPak processing eliminates local bottlenecks and speeds up end‑to‑end map creation.

• Cross‑platform performance: Multi‑frame localization on Meta Quest and Editor‑based simulation in Unity help teams iterate quickly and ship experiences that stay locked in place.

• Measured precision: MultiSet maintains 5 cm median accuracy with low drift in production‑grade deployments.

Developer links

• Unity Editor Localization Simulation → https://docs.multiset.ai/unity-sdk/localization-simulation

• NavVis scans (E57 w/ equirectangular panoramas) → https://docs.multiset.ai/basics/third-party-scans/navvis-scans

• Map Upload API (E57/third‑party scans) → https://docs.multiset.ai/basics/rest-api-docs/map-upload-api

• Matterport MatterPak on‑cloud workflow → https://docs.multiset.ai/basics/third-party-scans/matterport-e57/matterpak-files

• MatterPak Upload API (OAuth/credentials) → https://docs.multiset.ai/basics/rest-api-docs/map-upload-api/matterpak-upload-api

• Meta Quest SDK (samples, object tracking, multi‑frame) → https://github.com/MultiSet-AI/multiset-quest-sdk

• Xgrids scans (E57 export + ingest) → https://docs.multiset.ai/basics/third-party-scans/xgrids-scans

• Changelog → https://docs.multiset.ai/changelog

Availability

All updates in v1.9.1, v1.9.2, and v1.9.3 are available today across the MultiSet platform, SDKs, and Developer Portal.

About MultiSet AI

MultiSet AI is an enterprise visual positioning and spatial mapping platform that transforms physical spaces into precise, intelligent AR environments. The platform delivers centimeter‑level 6‑DoF localization with low drift and is designed for cross‑platform deployment across Unity, native iOS/Android, WebXR, Meta Quest, and ROS — enabling location‑aware apps at industrial scale.

Media contact

Shadnam Khan

Co-Founder

shadnam@multiset.ai