Roboflow

RF-DETR Large

About

RF-DETR Large is a 33.9M open-weight model from Roboflow. Open-vocab detection. It accepts text and images. Released May 11, 2026.

Parameters
33.9M
Published parameter count.
License
Apache-2.0
License on the weights.

Pricing

Token rates are US dollars per 1M tokens. Image rates are per 1K images. Audio rates are per 1M audio seconds.

Rate	Price
Input / 1M	—
Output / 1M	—

Benchmarks

Published scores for RF-DETR Large, from the RF-DETR docs.

Evals

Reading	Example
Qualitative	Clear structure, grounded in the input
Quantitative	COCO AP50: 75.1.
Cost and performance	Lower listed rate, mid-pack latency

Examples

Cupcake detection
RF-DETR nano draws boxes around every cupcake in 40 ms.
{"detections":[{"label":"cupcake","score":0.91,"box_xyxy":[114,236,302,418]}, ...]}

Methods

Methods this model serves. Payload shapes are in the docs.

Method	Returns
detection	Boxes around objects

Quick start

Call this model on the OpenAI-compatible gateway. The model id is already filled in.

from openai import OpenAI

client = OpenAI(
    base_url="https://gateway.vlm.run/v1/openai",
    api_key="<VLMRUN_API_KEY>",
)

response = client.chat.completions.create(
    model="roboflow/rf-detr-large",
    messages=[{"role": "user", "content": "What is in this image?"}],
)

print(response.choices[0].message.content)