#6565th of 70 in Open models
omdet
Real-time open-vocabulary object detector driven by text prompts
- Stars
- 1.4k
- License
- Apache-2.0
- Last commit
- Mar 2026
- Language
- Python
1,369 on 21 Apr 20261,410 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Official PyTorch implementation of OmDet-Turbo, a transformer-based open-vocabulary detector that finds objects from class names given as text. It uses an Efficient Fusion Head to reduce multimodal fusion cost. The repo provides a Tiny Swin-T model, a demo script, an API server on port 8000, and ONNX export, with zero-shot scores of 42.5 AP on COCO and 30.3 on LVIS.
Who it is for: Computer vision engineers needing text-prompted object detection
Strengths
- Tiny model reports 21.5 FPS in PyTorch and 140.0 FPS with TensorRT
- Detects arbitrary text-prompted classes zero-shot (COCO 42.5, LVIS 30.3 AP)
- Ships demo script, HTTP API server with /docs page, and ONNX export
- Integrated into Hugging Face Transformers 4.45.0; weights also on ModelScope
Weaknesses
- Model zoo lists only the Tiny Swin-T weights; Base weights are not linked
- ONNX export has fixed input sizes and no post-processing by default
- No training instructions, inference and export only
- Model and CLIP checkpoints must be downloaded manually into a resources folder
What it needs
- Needs PyTorch, CLIP ViT-B-16 checkpoint
- Models: OmDet-Turbo-Tiny (Swin-T), ONNX export
- port 8000
Quick answers
Is omdet open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Also in Open models
See all 70| Rank | Project | Score |
|---|---|---|
| 1 | ruviewWiFi CSI sensing on ESP32 for presence, breathing, heart rate and pose | 94 out of 100 |
| 2 | layaSingle-pass multilingual classifier for typed choice, score and yes/no decisions | 88 out of 100 |
| 3 | rf-detrReal-time transformer models for object detection, segmentation and keypoints | 79 out of 100 |
| #4 | timesfmPretrained time-series foundation model for zero-shot forecasting from Google Research | 78 out of 100 |
| #5 | yolov5PyTorch YOLOv5 object detection, segmentation and classification models with training code | 76 out of 100 |