What is an AI product video?
An AI product video is a short film of a real product generated from still photos rather than filmed on a set. Nitaris takes one photo of your product, builds a lit studio scene around it, and moves a virtual camera through that scene to produce a 10-second clip.
The distinction that matters is between generating a scene around a real product and generating the product itself. An image model asked for "a gold pendant on marble" will invent a plausible pendant, which is useless for commerce — the thing in the ad has to be the thing in the box. An AI product video in the sense Nitaris means it starts from your photograph and treats the product in it as fixed, then generates everything around it: the set, the lighting, the depth of field and the camera move.
That is also what separates it from the two things it gets confused with. It is not animation, because nothing is drawn or rigged. And it is not a slideshow — the pan-and-zoom effect a template app applies to a still photo — because the camera genuinely moves through a three-dimensional scene, so highlights travel across a surface and the background parallaxes behind the product. Those are the cues a viewer reads as "filmed" without being able to name them.
In practice the whole exchange is one photo in and a finished 10-second MP4 out, plus a set of still marketing photos from the same scene. There is no timeline to assemble, no footage to shoot and nothing to edit — which is the actual reason the category exists, rather than the technology behind it.
Answered?Now watch your own product.
One photo in — a cinematic 10-second video and a full set of marketing photos out. Your first one is free.