Turn a photo into video with AI

Upload your photo in the tool window and press the button. The AI decides for itself what in the frame can move: hair and fabric in the wind, water, leaves, a small change of expression and the eyes. If you already picture a particular movement, you can describe it — there is a field for that, but it is optional. It all happens in the browser: no editor, no plugins, no keyframes.

It is the same action named from the other end: one photograph goes in, a short clip comes out. There is no slideshow to assemble from several pictures, no music to pick and no timeline to cut — the movement appears inside the frame itself. The finished video can be downloaded and posted as it is.

A slideshow is several photographs shown one after another with transitions; the pictures themselves stay still. Here it is the other way round: there is one photo, and it is the photo that moves — the person, the fabric, the water, the background, the camera. That is why the clip looks filmed rather than assembled.

No. The tool runs from a single button, and the frame will come to life without a word from you. The note field is for people who already have the movement in mind: “a slow push in”, “a turn of the head”, “wind in the hair”. Write it in plain language, not as instructions for a model.

The point of animating is to keep the likeness: this is your photograph that started moving, not a new person. The features stay where they are. On small and murky photos a moving face can come out slightly different, simply because there is little readable detail in the file and some of it has to be built.

Frames where the subject is clearly visible: a face in close-up, a whole figure, a background you can read. Portraits, landscapes with water and sky, and animal photos all work well. It is harder where the person is badly blurred, hidden behind others or shot from far away: there is almost nothing to move, and the result would be invented rather than animated.