ultralytics/yolo-ios-app

Ultralytics YOLO iOS app and Swift package for real-time Core ML inference across major computer vision tasks.

What it solves

This project enables real-time, on-device AI vision tasks on iOS and iPadOS. It removes the need for cloud-based processing by providing a native framework to run YOLO models directly on Apple hardware, ensuring lower latency and better privacy.

How it works

The project consists of a main iOS application for immediate testing and a Swift Package (SDK) for developers to integrate into their own apps. It leverages Apple's first-party frameworks, specifically Core ML and the newer Core AI (.aimodel), to run inference on the Apple Neural Engine (ANE) and GPU. It supports a wide range of tasks including object detection, instance and semantic segmentation, depth estimation, classification, and pose estimation using YOLO11 and YOLO26 models.

Who it’s for

  • iOS Developers who want to add real-time computer vision capabilities to their apps using a lightweight Swift package.
  • AI Researchers and Hobbyists who want to test YOLO models on physical iPhones or iPads without writing a full app from scratch.

Highlights

  • High Performance: Achieves camera-rate inference (~30 FPS) on recent iPhones.
  • Broad Task Support: Handles everything from standard object detection to oriented bounding boxes (OBB) and depth estimation.
  • Zero Third-Party Dependencies: Built entirely with pure Swift and Apple's native frameworks.
  • Flexible Model Loading: Supports bundled nano models, on-demand downloads of larger variants, and custom user-provided Core ML/Core AI models.

Related

  • Project
    ultralytics/ultralytics

    A comprehensive computer vision framework featuring YOLO models for real-time object detection, segmentation, classification, and pose estimation.

  • Project
    FaceAISDK/FaceAISDK_iOS

    An on-device, offline SDK for iOS that provides face detection, recognition, liveness detection, and anti-spoofing capabilities.

  • Project
    ultralytics/yolo-flutter-app

    YOLO Flutter is Ultralytics' official Flutter plugin that enables on‑device YOLO inference (detection, segmentation, depth, classification, pose, OBB) for both iOS and Android. It provides a single Dart API—`YOLO` for single‑image inference and `YOLOView` for live‑camera inference—handling automatic download and caching of official LiteRT/Core ML models, custom model loading, metadata‑driven task inference, and optional Qualcomm NPU support.

  • Project
    Geekgineer/YOLOs-CPP

    A production-ready C++ inference engine that provides a unified API for the entire YOLO family, supporting tasks from object detection to metric depth estimation.

  • Project
    aparsoft/yolo-streamlit-detection-tracking

    YOLO Vision Studio is a Streamlit web app that runs Ultralytics YOLO models for detection, segmentation, pose estimation, text‑prompted object search, and multi‑object tracking. It shows live counts of objects in the current frame and unique IDs seen so far, with a churn metric to gauge tracking quality. Supports webcam, RTSP, YouTube, and local video sources, and offers four built‑in trackers with optional appearance ReID. Install via `uv sync` (or pip), run `streamlit run app.py`, and start analyzing videos instantly.