YOLOv8-S

yolov8Coming Soon

one-stage detector with CSPDarknet backbone

Parameters11.2M

GFLOPs28.6

Input Size640px

Best mAP44.2%

Licensenon-permissive

Architecture

Type

one-stage

Backbone

CSPDarknet

Neck

C2f-PAFPN

Head

Decoupled

Benchmark Results

Performance on COCO val2017 across different hardware configurations

Hardware	Runtime	mAP@50-95	FPS	Latency	VRAM
NVIDIA Jetson Orin Nano Super 8GB	ONNX Runtime FP32	44.2%	3.2	313.3ms	-
NVIDIA Jetson Orin Nano Super 8GB	PyTorch FP32	44.2%	22.1	45.1ms	79 MB
NVIDIA Jetson Orin Nano Super 8GB	TensorRT FP16	44.2%	39.7	25.2ms	14 MB
Raspberry Pi 5	HailoRT INT8	36.9%	68.0	14.7ms	-
Raspberry Pi 5	ncnn FP32	44.2%	5.7	175.5ms	-
Raspberry Pi 5	ONNX Runtime FP32	44.2%	2.4	414.2ms	-
Raspberry Pi 5	PyTorch FP32	44.2%	2.4	421.1ms	-

Speed Breakdown(NVIDIA Jetson Orin Nano Super 8GB)

6.2ms

291.6ms

9.5ms

Preprocess

Inference

Postprocess (NMS)

anchor-freenmsPaper: 44.9% mAP

Related Models (yolov8)

YOLOv8-N YOLOv8-M YOLOv8-L YOLOv8-X

Run any model with one line

LibreYOLO has the best catalogue of state-of-the-art detectors, all behind one MIT-licensed Python API.

from libreyolo import LibreYOLO, SAMPLE_IMAGE # LibreYOLO has the best catalogue of state-of-the-art models. model = LibreYOLO("LibreRFDETRl.pt") # RF-DETR-L (transformer flagship) results = model(SAMPLE_IMAGE, save=True) # run inference, save the annotated image # Swap in any other model, same one-line API (weights auto-download): # LibreYOLO("LibreYOLO9c.pt") # YOLO9-C # LibreYOLO("LibreYOLOXx.pt") # YOLOX-X # LibreYOLO("LibreDFINEx.pt") # D-FINE-X # LibreYOLO("LibreRTDETRr50.pt") # RT-DETR-R50