Pre-Trained Models¶

This page lists the built-in pre-trained models that ship with Roboflow Inference - popular open architectures trained on the COCO dataset and exposed under short aliases like rfdetr-small or yolov8n-640.

These aliases are convenience shortcuts: no API key is required to use them, and you don't have to know the underlying Roboflow project/version. Pick a model from the tables below and pass its alias as model_id.

Run a pre-trained model alias with either the HTTP client or the native package. See Run a Model for package installation.

from inference_sdk import InferenceHTTPClient

client = InferenceHTTPClient(
    # api_url="http://localhost:9001",  # for Self-hosted
    api_url="https://serverless.roboflow.com",
    api_key="ROBOFLOW_API_KEY",
)
results = client.infer(
    "https://media.roboflow.com/inference/people-walking.jpg",
    model_id="rfdetr-small",
)

from inference import get_model

model = get_model(model_id="rfdetr-small")

results = model.infer("https://media.roboflow.com/inference/people-walking.jpg")

Supported Pre-Trained Models¶

You can click the link associated with a model below to test the model in your browser, and use the ID with Inference to deploy the model to the edge.

RF-DETR¶

YOLO26¶

YOLOv11¶

YOLOv10¶

YOLOv8¶

YOLO-NAS¶