wkentaro/labelme

Health v2 · maintenance only

Adoption not part of the score

16130 stars · 3702 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Labelme is a graphical image annotation tool written in Python with a Qt interface, supporting polygon, rectangle, oriented rectangle, circle, line, and point shapes for creating labeled datasets. It also offers AI-assisted annotation with SAM2/SAM3 and YOLO-World models running fully offline, plus export to formats like VOC and YOLO.

Use cases

• annotate images with polygons for semantic segmentation
• draw bounding boxes for object detection datasets
• create instance segmentation training data
• label images for image classification
• AI-assisted click-to-segment annotation with SAM
• convert annotations to YOLO or VOC dataset formats
• annotate video frames for computer vision
• label keypoints with point annotations

When to choose

• you need a free, open-source, offline image labeling tool
• you are building segmentation or detection datasets in Python
• you want AI-assisted annotation without sending data to the cloud
• you need JSON annotations exportable to common ML formats

When to avoid

• you need collaborative, multi-user web-based annotation
• you want a no-install experience without Python (consider the paid standalone app)
• you need audio, text, or 3D annotation
• you require permissive licensing for redistribution (GPL-3.0)

Facets

application · maturity active

image-processing computer-vision machine-learning gui cli computer-vision image-processing machine-learning artificial-intelligence developer-tools python cross-platform windows image-annotation labeling segmentation bounding-boxes sam dataset-creation qt yolo voc polygon-annotation linux macos desktop

Member repositories

For agents

markdown · JSON · MCP: product_card(name="wkentaro/labelme")