Model weights · YOLOv9

Download YOLOv9 weights

Adds Programmable Gradient Information and the GELAN backbone to preserve detail through deep layers, reaching high accuracy with very few parameters. Pick a size below and download the official .pt checkpoint in one click, with published COCO accuracy, size and license all in view.

GPL-3.0Academia Sinica (Wang et al.) · 2024 · Original repo + Ultralytics (PyTorch)

YOLOv9 checkpoints, pick a size and download

Adds Programmable Gradient Information and the GELAN backbone to preserve detail through deep layers, reaching high accuracy with very few parameters.

  • YOLOv9t
    yolov9t.pt
    COCO mAP
    38.3
    Params
    2M
  • YOLOv9s
    yolov9s.pt
    COCO mAP
    46.8
    Params
    7.2M
  • YOLOv9m
    yolov9m.pt
    COCO mAP
    51.4
    Params
    20.1M
  • YOLOv9c
    yolov9c.pt
    COCO mAP
    53.0
    Params
    25.5M
  • YOLOv9e
    yolov9e.pt
    COCO mAP
    55.6
    Params
    58.1M

Scores are COCO mAP at 640px, published by the authors. Weights host: github.com. Clicking Download verifies the file and starts it straight from the official CDN.

How to load YOLOv9 weights

Install the Ultralytics package, then point the loader at the checkpoint. It downloads automatically on first use, or you can pass the local path to the file you downloaded above.

pip install ultralytics
from ultralytics import YOLO

# Downloads on first use, or pass a local path to your .pt file
model = YOLO("yolov9t.pt")
results = model("image.jpg")

Want to see the architecture? Export to ONNX and open it in the ONNX visualizer for a labelled diagram with shapes and parameter counts.

Open visualizer →

Choosing a YOLOv9 checkpoint

Read off the table above: what each step up the size ladder costs and returns for this family specifically, where it stops being worth it, and what the license actually permits.

What the size ladder buys

YOLOv9 spans YOLOv9t at 2.0M parameters to YOLOv9e at 58.1M, 29.1× the model, for 17.3 more COCO mAP (38.3 to 55.6). That works out to roughly 0.31 points per additional million parameters across the whole range, and the return is front-loaded: the early steps are much cheaper than the last one. Published figures from Academia Sinica (Wang et al.) at 640px, not re-measured here.

Where the ladder stops paying

The best value step is YOLOv9t → YOLOv9s, worth 1.63 COCO mAP points per million parameters. The worst is YOLOv9c → YOLOv9e at 0.08, or 20× less efficient, for 32.6M extra parameters and only 2.6 more points. If you are latency- or memory-bound, that is the step to skip; if you are accuracy-bound and the compute is free, it is the only place left to get it.

Where it sits in the catalog

Against the 8 families here that publish COCO mAP, YOLOv9's best checkpoint ranks 2nd at 55.6, against 57.5 for YOLO26. Read that as a tier indicator rather than a head-to-head: the families use different size ladders, input sizes and training recipes, and each number is the authors' own. The closest alternatives for object detection are YOLO26, YOLO11 and YOLO12.

Licensing, in practice

The weights are GPL-3.0. GPL-3.0 requires derivative works to be released under GPL too.

About YOLOv9

YOLOv9 (Academia Sinica, ECCV 2024) tackles the information bottleneck in deep networks with Programmable Gradient Information (PGI) and the Generalized ELAN (GELAN) architecture. It delivers high accuracy at low parameter counts and is integrated into Ultralytics for detection and segmentation.

Author
Academia Sinica (Wang et al.)
Released
2024
Tasks
Detect, Segment
Framework
Original repo + Ultralytics (PyTorch)
Input size
640px
License
GPL-3.0
Resources
DocsGitHubPaperLicense

More model weights to download

Each page lists every checkpoint with accuracy, parameters, license and a one-click download.

See all 19 model families →

YOLO tracker FPS benchmarks →Get in touch →
Blog

From the blog

Tutorials, code, and notes on computer vision, deep learning, and applied AI.

View all posts →

Frequently asked questions

How do I download YOLOv9 weights?
Click the Download button next to any variant in the table above; we verify the file and start it straight from the official CDN. You can also let the loader fetch it automatically on first use with `model = YOLO("yolov9t.pt")`. All 5 YOLOv9 checkpoints are hosted officially.
Is YOLOv9 free for commercial use?
YOLOv9 is released under GPL-3.0. GPL-3.0 requires derivative works to also be open-sourced under GPL; review it against your deployment. Always confirm against the linked license text.
Which YOLOv9 model size should I use?
Start with YOLOv9t, the smallest and fastest, ideal for prototyping, edge and CPU. Move up the ladder only when you need more accuracy and have the compute for it. The largest variant reaches 55.6 COCO mAP.