visionOS 27 expands ARKit object tracking with high-frame-rate tracking, metric-space pose queries, extended Create ML training, spatial accessory support, and brings object tracking to iOS for the first time.
⢠Added ReferenceObjectConfiguration API to enable high frame rate tracking per reference object
⢠Introduced ARKit Coordinate Space Correction API with 'rendered' and 'none' modes for metric vs display-corrected poses
⢠Extended training mode added to Create ML Object Tracking template for improved accuracy
⢠Object tracking (ARObjectAnchor, ARWorldTrackingConfiguration) now available on iOS 27 for the first time
⢠High-frequency tracking and extended training mode enable robust handheld object tracking for use cases like surgical training, home measurement, and assembly guidance
⢠The new ARKit Coordinate Space Correction API returns object poses in metric space (no display corrections), unlocking accurate spatial measurement applications
⢠Object tracking now works on iOS, meaning reference objects trained once in Create ML can be deployed to both iOS and visionOS apps
iOS 27 adds sectioned queries, codable model attributes, ResultsObserver for non-SwiftUI change observation, and HistoryObserver for reacting to persistent history changes in SwiftData.
iOS 27 expands the App Intents framework with RelevantEntities for contextual content surfacing, EntityCollection for high-performance batch parameter handling, SyncableEntity for cross-device Siri continuity, native Duration/PersonNameComponents parameter types, and @UnionValue for multi-type parameters ā plus extended intent execution time limits.
USDKit is a new first-party Swift framework introduced in iOS/macOS 27 that brings native USD scene creation, composition, modification, and export capabilities to Apple platform apps, with deep RealityKit and Spatial Preview integration.
In-depth guide
RealityKit & USDKit in iOS 27 āXcode 27 introduces an agentic localization workflow that lets you ask a coding agent to translate your entire app directly inside Xcode, leveraging String Catalog context ā including where and how strings are used ā to produce accurate, consistent translations across all languages.