1. Home
  2. 🛠️ Developer Tools
  3. Vision Detector
Vision Detector

Vision Detector

Run your Vision CoreML model

0 ratings
1 review
📊 Productivity
🛠️ Developer Tools
Free

Details

  • Released
  • Updated
  • October 9, 2022
  • July 10, 2026

Features

Developer

Kazufumi Suzuki
✨Promote your app ✨
Vision Detector screenshot #1 for iPhone
Vision Detector screenshot #2 for iPhone
iphone
ipad
⭐1
Reviews↘︎
🔄6+ apps
Alternatives↘︎
❓8 answered
FAQ↘︎
📦~ in 2 weeks
Updates frequency↘︎
🖼️Get Icon
Icons↘︎

About

Easily test and evaluate Core ML models on your iPhone or iPad. Load your own models from various cloud services or local storage and run them using live camera feed or stored images. Supports image classification, object detection, and style transfer.

✓Run Core ML models on device
✓Load models from cloud storage
✓Live camera feed inference
✓Image classification support
✓Object detection support
✓Style transfer support
Show original description

What's New in Vision Detector

3.1

July 10, 2026

Broadcasting Notification: Notification by LLM is now broadcasted to all your devices (iOS/macOS) via iCloudKit. Make a prompt to use Notification Center via tool calling: “Notify me using tool calling when a person is detected. Don't notify if there is no one.” Notifications are broadcast via iCloud to all iPhones, iPads, and Macs signed in with the same Apple Account. You can customize how notifications are displayed on each device in its system Notification settings. Notifications are delivered through Apple’s iCloud servers only. No other servers are used.

More

User reviews

Most RecentHighest RatedLowest Rated

Sadly it only supports webcam or live input.

I'm sure this is useful for many people and it is free (thank you dev), but for me it wasn't.

I wish we could use mp4 videos or even images to test our CoreML models quickly instead of being limited to webcam and live video inputs.

Anyway, it didn't seem right to rate it below 4-star as it is a free app.
Show more
BytufBytuf

Developer apps

AR Needle Guide
AR Needle Guide

AR guided procedure

🏥 Medical
Free
AR vProtractor
AR vProtractor

Virtual Protractor

📊 Productivity
Free
miniAssistant
miniAssistant

Tiny client for LM Studio API

🛠️ Developer Tools

(1)

Free
See developer apps

FAQ

What is Vision Detector?

Vision Detector is a free application that allows you to run Core ML models directly on your iPhone or iPad. It provides a simple interface for testing and evaluating machine learning models without needing to build a separate app in Xcode.

How do I load a model into Vision Detector?

You can load your Core ML model by copying it to your device using the Files app. Supported locations include iCloud Drive, OneDrive, Google Drive, Dropbox, or directly via AirDrop. Once the model is on your device, you can select it within Vision Detector.

What types of Core ML models does Vision Detector support?

Vision Detector supports image classification, object detection, and style transfer models. Models that require MultiArray inputs or outputs, or lack a non-maximum suppression layer, are not supported.

Can I use Vision Detector with live video?

Yes, Vision Detector supports live video input from the built-in camera. Inference runs continuously on the camera feed, with performance dependent on your device's capabilities.

Does Vision Detector come with pre-trained models?

No, Vision Detector does not include any pre-trained machine learning models. You must provide your own Core ML model in the correct format to use the application.

What devices is Vision Detector compatible with?

Vision Detector is compatible with Mac, iPhone, and iPad devices. It is designed to run on-device inference across Apple's ecosystem.

Is Vision Detector free to use?

Yes, Vision Detector is completely free to download and use. There are no in-app purchases or advertisements associated with the application.

How often is Vision Detector updated?

The latest version of Vision Detector is 2.2, last updated on February 2, 2026. While specific update frequency isn't detailed, this indicates ongoing development and support for the application.