LabelML
Label For Object Detection
Only for Mac
$2.99
Mac
Native macOS app for preparing Create ML object detection datasets with YOLO, COCO, and Pascal VOC export.
LabelML is a native macOS labeling tool for object detection datasets. It helps developers, machine learning engineers, and dataset maintainers create training annotations quickly and accurately.
Open a local image folder, draw bounding boxes on the canvas, move and resize selected boxes, delete mistakes, and manage classes with colors and visibility controls. The file sidebar helps you filter all, labeled, unlabeled, and issue images so you can focus on what still needs work.
LabelML supports familiar pro-app interactions: hold Space and drag to pan, use Command + mouse wheel to zoom, press A/D or arrow keys to move between images, use number keys to select classes, and press Command + S to save. For fine-tuning, the inspector lets you directly edit class, x, y, width, and height values.
LabelML saves annotations as the annotation.json file expected by Create ML object detector data sources. Before training, use the Create ML readiness check to see what already passes, what needs attention, and how to fix it. If extra JSON files are found in the folder root, LabelML can safely move them into a hidden backup folder so Create ML does not detect multiple annotation files. For other training pipelines, export your dataset to YOLO, COCO, or Pascal VOC.
Key features:
• Local image folder workflow with no account required
• Create, select, move, resize, and delete bounding boxes
• Add, rename, recolor, and delete classes
• Filter files by all, labeled, unlabeled, and issue states
• Precise annotation property editing
• Undo and redo for common editing actions
• Save Create ML object detector annotation.json by default
• Create ML training-readiness check and training-folder preparation
• Backup annotation.json before saving without interfering with Create ML
• Export to YOLO, COCO, and Pascal VOC
• Native macOS window, toolbar, and keyboard shortcuts
Ratings & Reviews
- This app hasn’t received enough ratings or reviews to display an overview.
• Fixed an issue where root-level annotation.backup.json or other extra .json files could be detected by Create ML and cause training import • failures
• Added a one-click Create ML training-readiness check showing satisfied items, issues, and suggested fixes
• Readiness results can prepare the training folder by safely moving extra JSON files into .labelml-backups/removed-json/
• Saving now produces the Create ML annotation.json directly, so a separate Create ML export step is not required
• Save backups are still preserved in a hidden backup folder without interfering with Create ML data sources
• Improved the Image Info panel with Create ML readiness details before training
The developer, 伟华 黄, indicated that the app’s privacy practices may include handling of data as described below. For more information, see the developer’s privacy policy .
Data Not Collected
The developer does not collect any data from this app.
Accessibility
The developer has not yet indicated which accessibility features this app supports. Learn More
Information
- Seller
- 伟华 黄
- Size
- 1.8 MB
- Category
- Utilities
- Compatibility
Requires macOS 12.0 or later.
- Mac
Requires macOS 12.0 or later.
- Mac
- Languages
English and 8 more
- English, French, German, Japanese, Korean, Portuguese, Simplified Chinese, Spanish, Traditional Chinese
- Age Rating
4+
- 4+
- Copyright
- © 2026 LiveSubX. All rights reserved.
