Home › Photo to Video › Photo to Video

How to turn a photo into a video with AI

Pick a photo in the Photo to Video app for iPhone and the AI generates a short video of it moving — a portrait that turns and blinks, a landscape whose water and clouds drift. A written prompt is optional: with no prompt the model infers plausible motion from the image itself, and with one you direct what moves. Ten models are available, clips run 1 to 20 seconds, and the highest resolution is 4K.

Steps

1. Open the app. Photo to Video opens on this mode — it is the home screen.
2. Pick a photo. Choose any image from your camera roll.
3. Optionally describe the motion. Leave it blank to let the model decide, or write what should move.
4. Choose a model, length and resolution. Ten models support this mode; see the table below.
5. Generate and save. Save the finished clip to your camera roll.

What it supports

Input	One image
Output	Video, 480p to 4K depending on model
Length	1–20 seconds depending on model
Models	Seedance 1.0 Pro (default), Seedance 2.0, Sora 2, Veo 3.1, Kling 3, Wan 2.7, Vidu Q3, Hailuo-02, Grok Imagine, Happy Horse
Prompt	Optional
Platform	iPhone (iOS). No Android or web version.
Pricing	Credits per second of output, by model. No subscription.

Which model to use

Seedance 1.0 Pro is the default and the cheapest per second — the right first try for any photo. Veo 3.1 is the one to pick when you need 4K or synchronised audio. Sora 2 and Seedance 2.0 handle complex scenes and camera movement most convincingly, at a higher cost per second. Hailuo-02 and Grok Imagine are quick, cheaper options for simple subject motion.

Getting a better result

• A clear single subject animates far better than a busy scene. Crop before you generate.
• Describe movement, not appearance. “Her hair lifts in the wind” gives the model something to do; “beautiful photo” does not.
• Start at 4–5 seconds. It costs a fraction of a 20-second clip and tells you whether the prompt works.
• If one model refuses an image, try another — each has its own content rules and switching is free.

What people use it for

A family photo you have only ever seen as a still

One image is all it needs, and the prompt is optional — leave it blank and the model infers breathing, blinking and small head movement on its own. Start on Seedance 1.0 Pro at 4 seconds; longer clips tend to drift on faces.

Prompt she breathes and blinks, small natural head movement, camera holds still

A portrait for a profile or a post

Subtle motion reads as intentional where a big movement reads as an effect. Keep it under 5 seconds and ask for one thing.

Prompt hair lifts slightly in a breeze, she turns her eyes toward the camera

A landscape or travel shot

Skies, water and foliage animate more convincingly than people, because small inaccuracies go unnoticed. This is where longer clips are worth the credits.

Prompt clouds drift left, water ripples, grass moves in the wind, slow push in

A product shot for a listing

Rotation and light movement sell an object without misrepresenting it. Veo 3.1 if the listing needs 4K.

Prompt the bottle rotates slowly on its axis, highlight sweeps across the glass

Where it struggles

Hands, teeth and any text in the image are where it breaks first. Fast or full-body movement from a single still is unreliable — that is what Motion Control is for. Group photos degrade faster than a single subject, because every face is another chance to go wrong.

Questions