# YOLO11-M

> YOLO11-M: one-stage object detector from the YOLO11 family. Planned for LibreYOLO (not yet packaged).

- Family: YOLO11 (External baseline)
- Parameters: 20.1M
- GFLOPs: 68.0
- Default input size: 640px
- Detection: anchor-free · nms
- Architecture: one-stage · CSPDarknet backbone
- Weight license: non-permissive
- Paper-reported mAP@50-95: 51.5%

## Benchmarks (COCO val2017)

| Hardware | Runtime | mAP@50-95 | FPS | Latency (ms) | VRAM (MB) |
|---|---|---|---|---|---|
| NVIDIA Jetson Orin Nano Super 8GB | ONNX Runtime FP32 | 50.6 | 1.5 | 686.2 | - |
| NVIDIA Jetson Orin Nano Super 8GB | PyTorch FP32 | 50.6 | 11.1 | 90.3 | 190 |
| NVIDIA Jetson Orin Nano Super 8GB | TensorRT FP16 | 50.6 | 25.8 | 38.8 | 14 |
| Raspberry Pi 5 | ncnn FP32 | 50.6 | 1.9 | 532.7 | - |
| Raspberry Pi 5 | ONNX Runtime FP32 | 50.6 | 1.0 | 994.5 | - |
| Raspberry Pi 5 | PyTorch FP32 | 50.6 | 0.9 | 1104.3 | - |


Source: https://www.visionanalysis.org/model/yolo11m. Benchmarks produced with LibreYOLO (https://github.com/Libre-YOLO/libreyolo).