Ultralytics YOLO26 is an open-source object-detection model that replaces the traditional Non-Maximum Suppression step with a dual-head architecture to output one box per object in a single pass. It is intended for researchers and enterprises needing fast, portable vision models and is distributed under AGPL for research with commercial licensing available.