# YOLO26-M

> YOLO26-M: one-stage object detector from the yolo26 family. Planned for LibreYOLO (not yet packaged).

- Family: yolo26
- Parameters: 20.4M
- GFLOPs: 68.2
- Default input size: 640px
- Detection: anchor-free · nms-free
- Architecture: one-stage · CSPDarknet backbone
- Weight license: non-permissive
- Paper-reported mAP@50-95: 53.1%

## Benchmarks (COCO val2017)

| Hardware | Runtime | mAP@50-95 | FPS | Latency (ms) | VRAM (MB) |
|---|---|---|---|---|---|
| NVIDIA RTX 5070 Ti | PyTorch FP32 | 52.2 | 84.1 | 11.9 | 191 |
| Raspberry Pi 5 | PyTorch FP32 | 56.2 | 0.9 | 1108.3 | - |


Source: https://www.visionanalysis.org/model/yolo26m. Benchmarks produced with LibreYOLO (https://github.com/Libre-YOLO/libreyolo).