YOLO26 is a unified real-time vision model family from Ultralytics designed to support a broad range of computer vision tasks, including object detection, instance and semantic segmentation, depth estimation, classification, pose estimation, and oriented object detection, with an emphasis on efficient end-to-end inference and deployment.

Original paper: Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models

YOLO26s

This model uses the YOLO26-Small (YOLO26s) variant for object detection, featuring a lightweight detection architecture with a default one-to-one end-to-end head that eliminates NMS during inference. It is well suited for real-time object detection on edge platforms and applications such as automotive perception, surveillance, robotics, and industrial video analytics.

Model Configuration:

• Reference implementation: Ultralytics forked by Ambarella
• Original Weight: YOLO26s_Weights.COCO2017
• Dataset: COCO
• Resolution: 3x640x640
• Support Cooper version:
  • Cooper SDK: [2.5.4]
  • Cooper Foundry: [2.3]

Model	Device	compression	Model Link
YOLO26s	N1-655	Activation_fp16	Model_Link
YOLO26s	X7	Activation_fp16	Model_Link
YOLO26s	CV7	Activation_fp16	Model_Link
YOLO26s	CV72	Activation_fp16	Model_Link
YOLO26s	CV75	Activation_fp16	Model_Link