Real-Time.
Segment Anything.
Track Everything.
On Device.
Real-time video object segmentation SDK for iOS. Tap once, track forever. No cloud needed.
Multi-target foot tracking
Tap-to-segment object tracking
Features
Built for Production
Everything you need to ship real-time video segmentation in your iOS app.
Segment Anything
People, body parts, pets, furniture, vehicles — any object you can tap. Powered by a foundation model that generalizes to arbitrary shapes, not just predefined categories.
Zero-shot, any objectReal-time Performance
30+ FPS on iPhone 14 Pro and later. Optimized from the ground up for Apple Silicon — smooth enough for live camera experiences.
30+ FPSTap to Track
User taps once on the target object. The SDK automatically tracks it across every subsequent frame — no bounding boxes, no additional annotation needed.
Point / box / mask promptsMulti-Target Tracking
Track multiple objects simultaneously with per-object mask output. Each target gets its own color-coded segmentation — perfect for comparative analysis.
Per-object masks100% On-Device
Runs entirely on the device. Zero data leaves the phone. Perfect for privacy-sensitive applications in health, fitness, sports analysis, and AR.
Zero cloud dependencyConstant Memory Usage
Stream hours of video without running out of memory. The SDK maintains a fixed memory footprint regardless of how long the video runs.
Stream foreverDrop-in Integration
Add SegmentKit via Swift Package Manager — one line in Xcode. Pre-built SwiftUI camera view included. Go from zero to real-time tracking in under 5 minutes.
1 line setupDeveloper Experience
Ship in Minutes, Not Months
SegmentKit handles the entire segmentation pipeline behind a clean Swift API. You get real-time video object tracking without worrying about model loading, memory management, or GPU scheduling.
import SegmentKit
"color:#6b7280;font-style:italic">// Initialize with your license key
SegmentKit.configure(licenseKey: "your-key")
let tracker = try ETKVideoTracker()
"color:#6b7280;font-style:italic">// First frame: user taps the target
let result = try tracker.startTracking(
frame: cameraPixelBuffer,
point: CGPoint(x: 0.45, y: 0.6)
)
"color:#6b7280;font-style:italic">// Subsequent frames: automatic tracking
let trackResult = try tracker.track(frame: nextFrame)
let maskOverlay = trackResult.overlayImage(on: originalImage)
"color:#6b7280;font-style:italic">// Access raw mask data
let mask: [Float] = trackResult.mask "color:#6b7280;font-style:italic">// 256×256 logits
let iou: Float = trackResult.confidence "color:#6b7280;font-style:italic">// IoU predictionUse Cases
Built for Any Domain
From fitness to AR to robotics — SegmentKit adapts to your use case with zero-shot segmentation of any object.
Sports & Fitness
Track body parts in real-time for form analysis, gait assessment, and motion capture.
Health & Rehabilitation
Monitor patient movements during exercises with per-limb segmentation and tracking.
Augmented Reality
Pixel-perfect object masks for AR overlays, background removal, and visual effects.
Robotics & Automation
Real-time object tracking for pick-and-place, navigation, and inspection tasks.
Pricing
Try Free for 7 Days, Then Choose Your Plan
Get full access to SegmentKit with a 7-day free trial. No credit card required — just your email.
Free Trial
Full access for 7 days — no credit card required
Indie
For indie developers and small studios
Pro
For teams building production apps
Enterprise
For large-scale deployments
7-day free trial with full access. All paid plans include SDK updates and bug fixes.
Payments processed securely by Paddle (Merchant of Record). Paddle handles all global tax compliance. Questions?
Documentation
Everything You Need to Ship
Comprehensive guides, API reference, and sample code to get you from zero to production.
Quick Start
Get up and running in 5 minutes. Add SegmentKit via SPM, configure your license key, and start tracking objects.
5 min readAPI Reference
Complete reference for ETKVideoTracker, ETKImageSegmenter, ETKTrackResult, and all configuration options.
Full referenceSample Projects
Full Xcode projects demonstrating video tracking, single-frame segmentation, and camera integration patterns.
3 examplesPerformance Guide
Tuning tips for different iPhone models. Understanding ANE vs GPU scheduling, memory budgets, and thermal management.
Advanced