roboflow/rf-detr
RF-DETR is a real-time object detection and segmentation model architecture developed by Roboflow, SOTA on COCO, designed for fine-tuning. [ICLR 2026]
Build with Backblaze B2
SDKs, agent skills, IDE extensions, and reference pipelines from Backblaze Labs. All open source.
Loading star history...
Where it stands
Rank, growth pace and release cadence relative to everything GStars tracks.
Ranks #523 of 2,845 repositories GStars tracks this week (top 18%).
5 releases in the last 90 days, the latest 9 days ago, with 112 contributors on GitHub.
Peaked at #18 in the weekly ranking the week of Oct 5, its best in the past year.
Use Cases & Benefits
- Implements a real-time transformer-based object detection and instance segmentation model architecture optimized for fine-tuning.
- Delivers state-of-the-art accuracy and speed on COCO and RF100-VL benchmarks, outperforming comparable real-time models in latency and precision.
- Use for developing fast, accurate object detection systems in applications requiring real-time performance on standard datasets.
- Use for fine-tuning segmentation models on custom datasets to achieve high-quality instance segmentation with low inference latency.
- Use for benchmarking and research in domain-adaptive object detection leveraging a modern transformer architecture with open-source code and pretrained weights.
Should you depend on it?
License terms, project hygiene and risks, from GitHub metadata and GStars' snapshots.
License
Apache 2.0
Permissive: commercial use, modification and redistribution are allowed with attribution; includes an explicit patent grant from contributors.
Community files
- README: present
- Contributing guide: present
Watch-outs
No watch-outs from GStars' checks.
General guidance from repository metadata, not legal advice. How we score
From the README
RF-DETR is a real-time, transformer-based object detection and instance segmentation model architecture developed by Roboflow and released under the Apache 2.0 license.
RF-DETR is the first real-time model to exceed 60 AP on the Microsoft COCO object detection benchmark alongside competitive performance at base sizes. It also achieves state-of-the-art performance on RF100-VL, an object detection benchmark that measures model domain adaptability to real world problems. RF-DETR is fastest and most accurate for its size when compared current real-time objection models.
On image segmentation, RF-DETR Seg (Preview) is 3x faster and more accurate than the largest YOLO when evaluated on the Microsoft COCO Segmentation benchmark, defining a new real-time state-of-the-art for the industry-standard benchmark in segmentation model evaluation.
RF-DETR achieves state-of-the-art performance on both the Microsoft COCO and the RF100-VL benchmarks.
The below tables shows how RF-DETR performs when validated on the Microsoft COCO benchmark for object detection and image segmentation.
Daily snapshots
Stars gained per day, and how forks and open issues moved, from GStars' daily snapshots
+1,242 stars over 90 days across 10 snapshot days; best day +28 on Sep 8.
Forks
1.2K+157
Open issues
82-18
Compared with peers in Computer Vision
Fastest-growing repositories GStars tracks in this cohort for the week of Oct 11, by 7-day star growth
| Repository | Stars | +7d | Health | Last release |
|---|---|---|---|---|
| koharu-rs/koharu | 6.4K | +483 | 63 | 30 days ago |
| PaddlePaddle/PaddleOCR | 91.0K | +0 | 75 | 123 days ago |
| microsoft/AI-For-Beginners | 69.8K | +0 | 45 | — |
| opendataloader-project/opendataloader-pdf | 29.7K | +0 | 75 | 2 days ago |
| memvid/memvid | 16.6K | +0 | 75 | 138 days ago |
| roboflow/rf-detr | 9.8K | +0 | 75 | 9 days ago |
| Category median | 29.7K | +0 | 75 | — |
Also by roboflow
Other repositories from this maintainer that GStars tracks
- Python51.3K stars
supervision
We write your reusable computer vision tools. 💜
- Python3.9K stars
trackers
Trackers gives you clean, modular re-implementations of leading multi-object tracking algorithms released under the permissive Apache 2.0 license. You combine them with any detection model you already use.
Discover Repositories
Search across tracked repositories by name or description