AI video, from one still image

Turn any photo into a dance video

Give Seedancer one photo — your dog, a plush toy, a cucumber, you — and pick a routine. It puts that subject through the whole dance, frame by frame, and hands you the video a few minutes later. Nothing to edit and nothing to film.

Tell me when it launches Coming to the App Store for iOS 16.0 and later.
  • 62 routines
  • 10 AI models
  • 1 photo needed
A cartoon penguin performing a full dance routine, generated from a single still image
A woman dancing head to toe in a gym, generated from one full-body photo
A bulldog boxing in a ring — a face-swap template that needs only a close-up photo

It does not have to be you

The subject needs a head and something like limbs. It does not need to be human, and most of what people make with this is not.

  • Your pet

    The single most convincing thing to animate, because nobody knows exactly how a beagle dances. One photo standing, taken at their level.

  • A character, a toy, a piece of fruit

    Plush toys, cartoon figures, a cucumber with a face on it. If it has a head and something like limbs, a routine can be mapped onto it.

  • You

    The hardest subject and the most impressive when it lands. One full-body photo in even light, arms clear of your sides.

How it works

Four decisions, and only one of them is a photo.

  1. 1

    Pick the movement first

    Browse 62 routines across five categories and choose the one you want. The template fixes the choreography, the framing and the length — everything except who is doing it.

  2. 2

    Add one photo

    One still image is the whole input. Head to toe for a dance template, or just a face for a Sports one. No footage of your own, and no second photo.

  3. 3

    Choose how good it needs to be

    Ten AI models, from the cheap fast ones you test on to the premium ones you finish on. Longer clips and better models cost more credits, and the price is shown before you commit.

  4. 4

    Close the app

    The generation runs on our servers, not on your phone, so leaving the screen does not cancel it. The finished video is waiting in My Videos, ready to save to your camera roll.

The longer version, including what makes a photo work

Five categories, 62 routines

Every one is recorded movement rather than a description of movement, which is why a template gives you the same routine every time.

A cartoon penguin performing a full dance routine, generated from a single still image

Viral Dance

19 routines · full-body photo

The routines people are already posting, timed to the beat. The largest category in the app and the one most first videos come out of.

A tuxedo cat in a pink outfit doing a simple dance, animated from a photo

Easy Fun

18 routines · full-body photo

Simple, forgiving motion. If a photo is going to work anywhere it works here, which makes this the category to test a new photo in.

A raccoon performing a long, technical dance routine

Dance Pro

11 routines · full-body photo

Long, technical routines. These run the longest and cost the most credits, and they are the ones that look choreographed rather than filmed.

A golden retriever puppy in a hoodie dancing, animated from a single photo

Impress Friends

10 routines · full-body photo

Big, unmistakable movement built for a group chat. Legible on a phone at arm's length, which is where these get watched.

A bulldog boxing in a ring — a face-swap template that needs only a close-up photo

Sports

4 routines · close-up photo

The only face-swap category. It puts a face into existing sports footage instead of animating a body, so a close-up head shot is all it needs.

All five categories

What is actually in the app

  • 62 routines, not a prompt box

    Every template is real recorded movement. Pick it twice and you get the same routine twice — which is the thing a written prompt cannot promise you.

  • Ten AI models

    Pixverse, Kling, Seedance and Veo, side by side in one picker. Test on the cheap ones, finish on the expensive one.

  • Face swap as well as motion

    The Sports templates put a face into existing footage instead of animating a body, so a close-up head shot is all they need.

  • It finishes without you

    Generation happens server-side. Lock the phone, switch apps, go and do something else — the video is in your library when it lands.

  • Or write your own

    Create Your Own takes a written description, an optional photo, a duration and a model, for the videos no template covers.

  • Screened before it renders

    Every upload passes a content check first. A refusal comes back straight away and costs nothing, because nothing was generated.

What it costs

Generating video is billed by the second, so the app is too. Credits make the two things that drive the cost visible at the point you choose them.

Two levers, both in front of you

  • Length, which is close to linear — half the clip is close to half the cost
  • Model, which is a large multiplier from the cheapest to the best
  • The cost of a generation is shown before you start it, never after
  • A failed generation is not meant to take credits

Two ways to hold them

Buy a pack of credits outright and keep them, or take a subscription that tops you up on a schedule. Packs suit people who generate in bursts; a subscription suits people who generate steadily.

Prices live in the App Store rather than on this page. They move with country and promotion, and a stale number here would be worse than no number at all.

Questions people ask first

Does it only work on photos of people?

No — and most of the library is not people. A pet, a plush toy, a cartoon character, a piece of fruit with a face drawn on it: if the photo shows one subject with something a model can read as a head, arms and legs, it can be animated. The dance templates want that subject shown head to toe.

What kind of photo works best?

For the dance templates, one subject alone in the frame, head to toe, upright, facing the camera, with nothing cropped at the ankles. For the Sports templates, a close-up of the face is enough. Even light and an uncluttered background do more for the result than a high-resolution camera does.

Does it need a video of me, or just a photo?

Just a photo. One still image is the whole input. The movement comes from the template you pick, not from any footage of your own.

How long does a video take to make?

Usually a few minutes. You can leave the screen or close the app while it runs — the generation finishes on our servers rather than on your phone, and the result lands in My Videos when it is done.

Can I write my own prompt instead of using a template?

Yes. Create Your Own takes a written description, an optional photo, a duration and a choice of AI model, and generates from that instead of from one of the 62 templates.

How do credits work?

Every generation costs credits, and longer clips and better models cost more of them. You can buy a pack of credits outright or take a subscription that tops you up on a schedule. Prices are shown in the App Store, because they vary by country and by promotion.

Where do my photos and videos go?

An uploaded photo is used to produce the video you asked for. Finished videos are yours to save to your camera roll or delete, and deleting your account removes your generation history. The full detail is in the privacy policy.

Can I animate a photo of someone else?

Only with their permission, and never a child. The automated screen catches unsafe content, not consent — whether the person in the photo agreed to be in the video is your call to make before you upload it, not something an app can check for you.

Longer answers about how AI video works

One photo is the whole input

Pick a routine, add a picture, and watch something that has never danced start dancing.

Tell me when it launches Coming to the App Store for iOS 16.0 and later.

Not on the App Store yet. One email when it is, and nothing else.