Labelme

Labelme is a desktop image annotation application for Windows, macOS and Linux that uses AI models running locally on the user's computer. It is made by the creator of the open source Labelme project, which began in 2016 and remains maintained. This listing covers the commercial application. Users open a folder of images and outline objects with polygons, bounding boxes or segmentation masks, then edit every point of the result. AI assistance comes from SAM2 and SAM3 and from YOLO World. A point or a rough box prompts SAM2, and typed text such as a plant description prompts SAM3 to annotate every matching instance. The app and each model are downloaded once and then work offline, so source images, annotations and exports stay on disk with no cloud upload, API key or third party AI account. Annotations can be checked for quality before export to YOLO and Pascal VOC formats for training. It is aimed at researchers, engineers and teams building computer vision datasets in fields such as agriculture, medical imaging, industrial inspection and traffic, who need to keep images private. Pricing is a one time purchase with no subscription, shown in US dollars. Starter is $49 and includes the app with one click SAM2 annotation and SAM3 text prompts. Pro is $79 and adds a dataset toolkit of more than ten terminal tools that require Python, ready to train exports, guides and priority email support. Pro Lifetime is $249 with free updates forever. Starter and Pro include one year of updates, and the purchased version can be kept afterward. A free three day trial and a 14 day money back guarantee are offered.