Pyyan / Compare / YOLO-World vs RF-DETR vs YOLO26 vs RTMDet
YOLO-World vs RF-DETR vs YOLO26 vs RTMDet
4 of 5Object Detection · verified 13 Aug 2026
| Specification | YOLO-World | RF-DETR | YOLO26 | RTMDet |
|---|---|---|---|---|
| Summary | Open-vocabulary detection at YOLO speed. | First real-time detector past 60 mAP on COCO. | The latest in the line most people mean by object detection. | Pure throughput, and an MIT licence. |
| COCO mAP | ~46 zero-shot | 60.5 | ~55 | ~52 |
| Speed | High | Real time | Very high | 300+ |
| Licence | GPL-3.0 | Apache 2.0 | AGPL-3.0 | MIT |
| Family | Open vocabulary | Transformer, set prediction | Single stage | Single stage |
| Open vocabulary | Yes | No | No | No |
| Released | 2024 | 2025 | 2026 | 2022 |
| Category | Object Detection | Object Detection | Object Detection | Object Detection |
| Official | Tencent ↗ | Roboflow ↗ | Ultralytics ↗ | OpenMMLab ↗ |
Highlighted rows are where these differ.
YOLO-World
- The speed of YOLO with text-defined classes
- Weaker than Grounding DINO on hard prompts
Best for prompted detection that has to be fast.
RF-DETR
- Leads the RF100-VL domain-transfer benchmark
- DINOv2 backbone, markedly better on occluded objects
- Apache 2.0, which YOLO is not
Best for most custom detection tasks.
YOLO26
- Strong on Jetson, Snapdragon and ARM CPUs
- AGPL-3.0: open-source your derivative work or buy a licence
Best for edge and mobile deployment.
RTMDet
- Wins on raw speed where accuracy is sufficient
- MIT, so no derivative-work obligation
Best for high frame-rate video.