SAM 3 is a foundation model for promptable segmentation
Analyze images to detect objects, points, keypoints, or text