# lyuwenyu/RT-DETR

[CVPR 2024] Official RT-DETR (RTDETR paddle pytorch), Real-Time DEtection TRansformer, DETRs Beat YOLOs on Real-time Object Detection. 🔥 🔥 🔥

Repository: https://github.com/lyuwenyu/RT-DETR
Canonical: https://ross.abutalabs.com/products/rt-detr
Language: Python
License: Apache-2.0
License Family: permissive
Topics: rtdetr, rtdetrv2
Last push: 2026-08-17T11:54:36+00:00

## Health v2 (maintenance only)
Score: 74/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 98, release rhythm 35, longevity 86
- inputs: {"age_days": 1211, "days_push": 16, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5476, forks 648 (observed 2026-08-28T04:09:19.655159+00:00)

## What it is
Official implementation of RT-DETR and RT-DETRv2, real-time object detection transformers that outperform YOLO models, in PyTorch and PaddlePaddle. It provides pretrained model weights, training/inference tools, and torch hub support.

## Use cases
- real-time object detection with transformers
- train a DETR-based detector on custom dataset
- find a YOLO alternative for object detection
- run pretrained RT-DETR inference via torch hub
- benchmark real-time detection model accuracy and speed
- detect small objects with sliced inference

## When to choose
- you need state-of-the-art real-time object detection with transformer architecture
- you want official pretrained weights and reproducible CVPR paper results
- you prefer DETR-style end-to-end detection without NMS post-processing

## When to avoid
- you need a lightweight detector for edge devices with minimal dependencies
- you only need simple image classification rather than detection
- you require a mature ecosystem with extensive plugin support like Ultralytics YOLO

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, machine-learning, artificial-intelligence
- platform: python, cross-platform
- tags: object-detection, detr, transformer, real-time-inference, pytorch, paddlepaddle, cvpr-2024, yolo-alternative, model-zoo, gpu

## Member repositories
- lyuwenyu/RT-DETR (main) score 74

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:19.655159+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:56:42.721311+00:00, confidence not recorded.
  - readme: https://github.com/lyuwenyu/RT-DETR (fetched 2026-08-28T04:09:19.655159+00:00, sha 7044bec59192)
- Data as of 2026-08-30T08:39:29.467469+00:00.