How to pick a pet photo for an AI video

1. One pet per photo

One photo makes one video, and one pet is the right unit. If you have two pets, make two videos, each from its own photo — a photo of both animals put together comes out muddled.

2. Show the whole animal, or at least the head and shoulders

The photos that work best show your pet from nose to tail, standing or sitting, looking towards the camera.

A head-and-shoulders shot still works if that is all you have. But the dance uses the whole animal, so legs and tail in the picture give a truer result.

3. Keep the light even, and skip the extras

Daylight near a window is enough. Avoid dark rooms, on-camera flash and hard shadows across the face.

No sunglasses, no hats pulled over the eyes, no beauty filters. Anything that changes how your pet looks will show up in the clip, because the clip follows the photo.

4. Upload, then look at the free still

Drop the photo into the generator and tick the box to confirm it is your pet or that you have permission to use it. The still comes back in under a minute.

The still is free, up to 2 per account, and it is a real frame of your video, so what you see is what the clip starts from. If it looks wrong, try a different photo before paying anything.

5. Pay, then let it render

Pick a pack — a single video costs $5.99, and packs bring the price per video down — and the render starts.

Most videos land in about 3 minutes. You can close the tab, because we email you when it is done. If a render fails, the credits are refunded automatically and retrying costs nothing.

6. Download it, then post it

When it is ready, download the MP4 from the page or from My videos. It is 6 seconds, 9:16, with sound and no watermark.

Post it wherever you like, and keep the AI-generated label on it. Your photo is deleted within 24 hours of upload either way.