Object Detection & Segmentation

AI models for detecting, localizing, and segmenting objects in images and videos — enabling precise understanding of visual content at a pixel level. This collection features the SAM (Segment Anything Model) family — SAM3 Image, SAM3 Video, SAM V2 Image, SAM V2 Video, SAM V2.1 Hiera Large — as well as SAM 3D Objects and SAM 3D Body for three-dimensional segmentation. SAM models from Meta AI can segment any object in any image with remarkable accuracy using point prompts, bounding box inputs, or fully automatic segmentation. These models are foundational tools for computer vision applications: automated image annotation and dataset creation, product isolation for e-commerce catalogs, medical image analysis, satellite imagery analysis, video object tracking, and any workflow that requires understanding what objects exist where in an image. SAM3 (Segment Anything Model 3) extends capabilities to video, maintaining consistent object segmentation across frames — critical for video editing automation and tracking applications. On Segmind, all detection and segmentation models are available as pay-per-use APIs — pass an image or video and receive precise segmentation masks in return. Chain with background removal or 3D creation models in Segmind Workflows for automated visual extraction pipelines.