Upload an image or video, and Video-LLaVa will give you a text description of what it "sees."
This model is cold. You'll get a fast response if the model is warm and already running, and a slower response if the model is cold and starting up.