Pyyan / Compare / OWLv2 vs RTMDet
OWLv2 vs RTMDet
2 of 5Object Detection · verified 13 Aug 2026
OWLv2Googlecurrent
RTMDetOpenMMLabcurrent| Specification | OWLv2 | RTMDet |
|---|---|---|
| Summary | Open-vocabulary detection trained with self-labelling. | Pure throughput, and an MIT licence. |
| COCO mAP | ~45 zero-shot | ~52 |
| Speed | Moderate | 300+ |
| Licence | Apache 2.0 | MIT |
| Family | Open vocabulary | Single stage |
| Open vocabulary | Yes | No |
| Released | 2023 | 2022 |
| Category | Object Detection | Object Detection |
| Official | Google ↗ | OpenMMLab ↗ |
Highlighted rows are where these differ.
OWLv2
- Accepts an image as the query, not just text
Best for querying by example image.
RTMDet
- Wins on raw speed where accuracy is sufficient
- MIT, so no derivative-work obligation
Best for high frame-rate video.