Upload an image, select detectors, and click Run Inference. Captions beneath each result show which model (and latency) generated it.