A powerful computer-vision annotation tool designed for fast dataset creation. Build training data for object detection and segmentation easily.
Create bounding boxes, polygons, and segmentation masks with a smooth UI.
Drag & drop or upload images instantly.
Organize projects, revisions, and label sets efficiently.
Optional AI-assisted labeling to speed up your workflow.
Supports YOLO, COCO, VOC, KITTI, and custom output formats.
No installation needed. Runs fully inside your browser.
Images stay on your device — nothing is uploaded to servers.
HexLabel is completely open-source and created for developers, students, and beginners in computer vision. Build training datasets for your projects quickly and efficiently — without complex software.
I develop free tools for the computer vision community.
You can contribute by submitting code on GitHub or sharing new ideas.
If my work helped you, buying me a coffee helps cover server costs.
Source Code
Buy Me a Coffee