How to Turn a Photo into a Video with AI

Short answer

To turn a photo into a video, upload the sharpest version of the image, choose one camera move (push in, pan, parallax or orbit), set a length of 5 to 15 seconds and a format such as 9:16 at 1080 x 1920, and ask for the image to stay exactly as shown. Add text and sound last.

Product launch

Real outputs from the Videotto template gallery.

A worked photo to video example

Example: one mug photo, one 10 second clip

This is an illustration of the process, not a customer story. The input is a single product photo: a ceramic mug on a linen cloth, shot square at 3000 x 3000 pixels, with soft window light from the left. The goal is a 10 second vertical clip for Reels and TikTok.

The request was four lines long: turn this into a 10 second 9:16 video, keep the mug exactly as shown, slow push in with gentle depth, then the name and price slide in from the left and hold. Soft acoustic music, a light tick when the text lands.

What the 10 seconds look like

Time	What happens on screen	Sound
0 to 4 s	Slow push in from the full frame to roughly 120%, the background drifts a little slower than the mug	Music fades in
4 to 5 s	A soft band of light moves across the glaze	Music only
5 to 7 s	Stoneware Mug, 350 ml slides in from the left, the price fades in under it	A light tick on each line
7 to 10 s	Motion slows to almost nothing, text and logo hold still	Music resolves

Before posting, the result was checked against the original photo side by side: the handle shape, the glaze colour and the cloth pattern all had to match. If anything had drifted, the fix would be one more line in the prompt, "do not change the product", and a second render.

Why this photo to video edit works

Four choices carry most of the result, and they apply to any photo, not only products.

One camera move. The clip has a single push in. Zoom, spin and shake together read as cheap, while one clear move at a slow speed reads as deliberate.
A big enough source image. A 1080 x 1920 frame pushed in to 120% needs about 1,300 pixels across the visible area to stay sharp. Start from the largest file you have, ideally 2,000 pixels or more on the short side.
Text that holds long enough. A rough rule is one second for every three words, plus one more. Six words of copy need about three seconds of hold before the clip ends.
Faithful to the original. Asking for the image to stay exactly as shown stops the motion inventing details. For anything you sell or build, a gentle move on the real thing beats a dramatic move on something that looks different.

Under 5 seconds the move barely registers. Past 15 seconds one image starts to feel stretched.

Which motion suits which photo

In our study of 1M+ videos by 280,000+ users on Videotto, "animate this image" is one of the most common ways people start a project with no footage. The type of image decides the best move.

Image	Move that suits it	Typical length	How people ask for it
Product photo	Push in or slow orbit, details called out	6 to 10 s	"make the product float and spin, details appearing from the sides"
House or interior render	Smooth glide through the space	8 to 15 s	"as if filmed by an architectural videographer"
Illustration or sketch	Characters or parts that move	5 to 8 s	"have the character walk from left to right"
Poster or flyer	Text and shapes animate in, layout kept	6 to 12 s	"animate this exact design, keep the layout"
Lower third or logo	Clean reveal, then hold	3 to 5 s	"animate this lower third on a transparent background"
Old photo	Gentle zoom with slight depth	8 to 12 s	"bring this photo to life with a slow zoom"

If you are unsure, use a push in. It suits almost every image and is the hardest move to get wrong. Pull outs work for reveals, pans for wide landscapes and interiors, parallax for photos with a clear subject against a separate background, and orbits for single objects on a plain backdrop.

Read: How to Turn a Product Photo into a Promo Video

When the photo is something you sell and the clip needs to sell it.

A photo to video prompt you can paste

Adapted from real requests Videotto users sent to Otto. Fill in the brackets and say the context first: "this is a render of a house exterior" or "this is my product, a ceramic mug" helps the motion fit.

Paste into Otto

[upload your image]
This is [what the image is].
Turn it into a 10 second 9:16 video.
Keep the image exactly as shown: same design, colours and composition. Do not add new objects.
Motion: one slow push in with gentle depth, then [the text or detail] animates in and holds for 3 seconds.
Mood: [calm and premium / bright and fun / cinematic].
Add [soft music / ambient sound / no sound].

Photo to video tools compared

There are several ways to animate a still. The time figures are typical estimates for one 10 second clip by someone comfortable with the tool, not measured results.

Method	What you do	Typical time	Good for
CapCut	Add the photo, set keyframes for scale and position, add text and music by hand	10 to 20 minutes	Simple push ins and pans on your phone
Canva	Apply a page animation and text animations to a design	5 to 15 minutes	Posters and social graphics with preset motion
After Effects with Photoshop	Cut the subject out, rebuild the background, set up 3D layers and a camera for parallax	1 to 2 hours	Full control over depth and timing
Otto in Videotto	Upload the photo, describe the move, text and sound, check the result	About 5 minutes of writing and review, plus render time	Getting a finished clip with motion, text and sound from one request

Keyframing in CapCut is fine for one clip. The gap grows when you need parallax, several versions or a batch of images in the same style, which is where describing the result is quicker than building it.

Read: The Complete Guide to Making Videos from Scratch with AI (No Footage)

When one image is not enough and you need a full video.

FAQ

How long should a video made from one photo be?

A video made from one photo is usually 5 to 15 seconds long. That is enough for one camera move and a line of text to land. If you need 30 seconds or more, use three to six images in sequence, or mix the photo with text cards and other scenes, so the viewer has something new every few seconds.

What resolution does a photo need to be to turn it into a video?

For a 1080 x 1920 vertical video, a photo should be at least 1080 pixels on its short side, and closer to 2,000 pixels if the camera pushes in. Each step of zoom crops into the image, so a larger file keeps the final frames sharp. Phone photos from the last few years are usually big enough.

How do you make a still photo look 3D?

A still photo looks 3D when the subject and the background move at different speeds, which is called a parallax effect. Done by hand, it means cutting the subject out, filling the gap behind it and animating the layers separately. Photos with a clear subject and some distance to the background give the most convincing depth.

Can you turn an old family photo into a moving video?

You can turn an old family photo into a moving video with a slow zoom and slight depth, which suits archive images well. Scan the print at 600 dpi or more first so there is enough detail to move into. Keep the people unchanged. Gentle camera motion tends to look more respectful than animating faces.

Do I need permission to animate a photo I found online?

You do need permission to animate a photo you found online, unless its licence already allows it. A photo is protected by copyright whether it is still or moving, and adding motion does not change who owns it. Use your own images, licensed stock, or photos where the owner has given you permission.

Try it on your own video

Make your next video with Otto

Tell Otto what you want in plain words. It cuts, captions and animates the video for you.

Try Videotto Watch how it works