Make a POV video from the viewer's eyes

A POV video works when the viewer forgets there is a camera at all. The lens stands in for their eyes, so it stays at the height and distance a real person would have: seated, standing or lying back, never floating above the room. The person in the scene faces the lens and reacts to it as if it were someone, not an object.

Three details carry most of the effect. First, a fixed vantage point: the camera does not cut to a wide shot or circle the room. Second, framing that matches human sight, usually a close-up or a medium shot. Third, eye contact. When your partner looks straight into the lens, the viewer feels looked at, and that is what separates POV from an ordinary clip.

Start the prompt with the camera, then the person. Say where the viewer is (“first-person view from an armchair”), the angle (eye level, or a low angle looking up) and the shot size (close-up, medium shot). Then describe your partner: where they stand or sit, what they wear, the light in the room and where they look.

A complete prompt can be short: “First-person view from a sofa, eye level, medium close-up. A woman with short dark hair sits facing the camera, looks straight into the lens and smiles. Warm lamp light, steady camera, no change of shot.” For a video longer than 5 seconds, write the scenes in order and repeat the same vantage-point line in each one. Nureta stitches the scenes into one video, so every scene starts from the same place.

Yes, with image-to-video. You can add up to 2 reference photos, and they must show consenting adults. In POV the camera is the viewer, so the photo usually shows the person who faces the lens, for example a partner who agreed to take part. If you want to appear yourself, use your own photo: the scene then shows you facing the camera instead of looking through your eyes. Text-to-video needs no photo at all; a written description of the person is enough.

The content policy bans real people without their consent and public figures. Reference photos that look like a celebrity or public figure are rejected automatically, so a look-alike photo will not go through. Nureta is strictly 18+, and you confirm your age before you generate. Every video you make is private by default.

Nureta has about 200 ready-made scenes in about 30 categories, and one of them is POV. Every scene in that category is built around a first-person camera: the vantage point, framing and length are already set, so you can generate without writing a prompt. Open the scenes page, choose the POV category and pick the setup that fits.

If none fits, write a fully custom scene with the same rules the ready-made POV scenes follow: one fixed vantage point, a close or medium shot, and a partner who faces the lens. A ready-made scene is the fastest way to see what a 5-second POV clip looks like, and a custom scene gives you full control over the room, the light and the person.