ONLINEAGENT_OPS 2026.Q3 HOME HUBSMODELSBUILDINGCOMPANIESFILMMAKINGSECURITYLIVING WITH ITSKILLSALL HUBS, ONE TERMINAL → ARTICLES CRAFT RECORD BLOG TOOLS SKILLS SEARCH /
HOME›ARTICLES›HOW TO TURN A PHOTO INTO A VIDEO WITH AI
ARTICLES · HOW-TOBEGINNER

How to Turn a Photo Into a Video With AI

Turn a photo into a short AI video in five steps: use a sharp photo you may use, describe only the motion, get consent for real people, then check it.

READ10 min
WORDS1,901
SECTIONS8
TYPEHOW-TO
CHECKED1 OCT 26

You can turn a photo into a short video with AI in about five steps: pick a sharp photo you have the right to use, upload it to a video tool, describe only the movement you want, generate a few versions, and check the result before you share it.

If the photo shows a real person, get their permission first. A moving picture of someone is much easier to mistake for the real thing than a still one.

TL;DR — THE SHORT VERSION
  • How it works: the photo becomes the first frame. The AI imagines what happens next, a few seconds at a time.
  • Describe the motion, not the photo. The tool can already see the picture. Tell it what moves and how the camera moves.
  • One movement per clip. A small, simple action survives. Lots of action at once tends to fall apart.
  • Real people need consent. Only animate someone’s photo with their permission, and never in a way that could fool or embarrass them.
  • Label it. If you post it, say it was made with AI.

What you need

The plain point: a good photo, permission to use it, and a video tool that accepts an image.

  • A sharp, well-lit photo. Runway, which makes video AI, says the starting image should be high quality and free of flaws.1
  • The right to use it. Your own photo of a view, a pet or an object is simplest. A photo of a person needs their permission (see the next section).
  • A video tool that takes a photo. Google’s Gemini app and Runway are two examples, and the site’s cheat sheets cover others. Most need an account, and some need a paid plan. Veo vs Gemini Omni vs Grok Imagine Video covers what the companies publish, and this page gives no prices.
  • An idea of one movement. “The waves roll in.” “She turns and smiles.” “The camera slowly moves closer.”

The comparison: it is like handing a film crew one photo and asking, “What happens in the next five seconds?” They will keep the scene, but every movement you do not describe, they have to invent.

A PHOTO INTO A VIDEO · FIVE STEPS
The photo becomes the first frame. Your words say what happens next.
Pick a sharp photo you may useA real person in it? Ask them first.
Upload it to a video toolOne that accepts a photo to start from.
Describe only the movementWhat moves, how the camera moves, and how fast.
Generate a few versionsChange one word at a time. Keep the best.
Check it before you shareFaces, hands and text. Say it was made with AI.
Reasoning, 1 Oct 2026 — summarises this page’s opening answer and its sections on consent, the steps and checking the result.

The plain point: animating a photo of a real person makes them appear to do something they never did. Ask them first, and never make something that could deceive or hurt them.

Google’s own help page for making videos from photos says to upload only photos you have the right to use.2 Its rules for all its AI tools also ban pretending to be a real person in order to deceive.3

  • Ask the person. “Can I animate this photo of you and post it?” is a short message. A clear yes is the line between a fun clip and a problem.
  • Children: ask a parent, and think twice before posting at all.
  • Someone who has died: ask the family. An animated old photo can be moving for one relative and upsetting for another.
  • Famous people and strangers: do not. The law on using a person’s face or voice differs by country and, in the US, by state. Cloning a voice or a face explains the US position.
◈ IN PLAIN TERMS

Owning the photo is not the same as having permission to make the person in it move. Ask: would they be happy to see this clip, and would a stranger know it is not real? If either answer is no, do not make it.

The steps

The plain point: upload the photo, describe the movement, generate a few versions, pick the best.

  1. Prepare the photo. Crop it to the shape you want the video to be (tall for phones, wide for screens). Aspect ratio, resolution and format explains the shapes.
  2. Open the tool’s video maker and add your photo. In the Gemini app on a computer, Google’s steps are: open the sidebar, choose Create video, type a prompt, click Add image to upload your photo, then Submit.2 Other tools have a similar “image to video” or “add image” option.
  3. Describe the motion only. Runway says the photo already supplies the scene, so the prompt should describe movement.1 Say what moves, how the camera moves, and how fast.
  4. Start small. Ask for one simple movement first. Add more only once that works.4
  5. Generate a few versions and keep the best one. Each try comes out a little different, so change one word at a time to learn what helps.
TAKEAWAY

The photo is the set and the actors. Your prompt is the director shouting “action”: say what happens, not what is already there.

A starter prompt

The plain point: camera first, then the subject’s one action, then the mood and speed.

PHOTO TO VIDEO — ADAPT THE BRACKETS
The camera [slowly pushes in] as [the dog lifts its head and wags its tail]. [Leaves drift gently in the background.] Keep the dog’s face, markings and the setting exactly as in the photo. Calm, natural movement, [real-time speed], no sudden changes.

This follows the order Runway recommends for image-to-video prompts.1 Other tools prefer other orders, so check the cheat sheet for your tool. Camera grammar lists the camera moves worth naming.

Common mistakes

The plain point: most bad clips ask for too much, or start from a weak photo.

  • Re-describing the photo. It wastes the prompt and can confuse the tool. Describe the movement.
  • Too much at once. A moving camera, a moving person and a changing background together usually falls apart, as AI images and video for beginners explains. Pick one.
  • A blurry or busy photo. It gives the tool less to work from. Runway asks for a clean, high-quality starting image for this reason.
  • Expecting the same face every time. Each version is made fresh, so a face can change. Character consistency and how to use reference images explain the fixes.
  • Skipping consent because it is “just for fun”. Clips travel further than you expect.

When a clip goes wrong, why your generation came out wrong matches common failures to their causes.

Check the result

The plain point: watch it all the way through, twice, before you share it.

  • Faces and hands: does the person still look like themselves in every second? Are there extra fingers or a changing face?
  • Words in the picture: check that signs and labels still read correctly. Getting text right inside an image explains why text is fragile.
  • The start and the end: watch the first and last second closely, and trim anything odd.
  • Could it mislead? Would someone think this really happened? If so, add a clear note that it was made with AI.
  • Labels: some tools add a hidden mark, but many do not, and a missing mark proves nothing. How AI content gets labelled explains this. Your own label in the caption is the one that cannot be stripped.
◈ IN PLAIN TERMS

If a stranger could believe the clip is real, say it is not. A one-line caption, “animated with AI from a photo”, protects both you and the person in it.

Where to go next

Questions people ask

What should I write in an image-to-video prompt?

Describe the motion, not the photo. Runway says the image already supplies the scene, so the prompt should describe movement.1 Say what moves, how the camera moves, and how fast.

How do I turn a photo into a video in Gemini?

On a computer, open the sidebar, choose Create video, type a prompt, click Add image to upload your photo, then Submit.2

Can I animate a photo of someone else with AI?

Only with their permission. Google’s help page for making videos from photos says to upload only photos you have the right to use.2 Do not make anything that could fool or embarrass the person, and do not animate famous people or strangers.

Why does my AI video from a photo fall apart?

Usually the prompt asks for too much at once, or the photo is weak. Start with one simple movement, and use a sharp, well-lit photo: Runway says the starting image should be high quality, with no visual flaws.1

QUICK CHECK · 3 QUESTIONS

Pick an answer to see if you got it. Nothing is recorded.

1. What does the page say your prompt should describe?

  • Everything in the photo, so the tool understands the scene
  • The movement: what moves, how the camera moves and how fast
  • The name of a film whose style you want
Show the answer

Answer: The movement: what moves, how the camera moves and how fast. The page says the tool can already see the picture, so re-describing it wastes the prompt; describe only the motion.

2. How much action does the page suggest asking for in a clip at first?

  • A moving camera, a moving person and a changing background together
  • As much as possible, then trim the result
  • One simple movement
Show the answer

Answer: One simple movement. The page says a small, simple action survives, while lots of action at once tends to fall apart.

3. What does the page say about animating a photo of a real person?

  • Ask them first, because owning the photo is not permission to make them move
  • Owning the photo is enough permission
  • It is fine without asking if the clip is just for fun
Show the answer

Answer: Ask them first, because owning the photo is not permission to make them move. The page says an animated photo makes someone appear to do something they never did, so get a clear yes and never make something that could deceive or hurt them.

SOURCES
  1. 1Runway Academy, Prompting Guide, undated, read at source 1 Oct 2026“For best results, ensure that the input image is high quality and free of visual artifacts.” ↑a“Your input image acts as the first frame and provides the model with the composition, subject matter, lighting, and style information for the video.” and “Rather than describing elements present in the image, use your prompt to describe the motion of the scene.” ↑b“For Image to Video: The camera [motion description] as the subject [action]. [Additional descriptions]”. ↑c“Your input image acts as the first frame and provides the model with the composition, subject matter, lighting, and style information for the video.” and “Rather than describing elements present in the image, use your prompt to describe the motion of the scene.” ↑d“For best results, ensure that the input image is high quality and free of visual artifacts.” ↑e
  2. 2Google, Gemini Apps Help, Generate videos with Gemini Apps, undated, read at source 1 Oct 2026“Only upload photos you have the right to use”. ↑asteps include “Create video”, “Add image” and “Submit”; Google also says “This feature isn’t available to users under 18”. ↑bsteps include “Create video”, “Add image” and “Submit”; Google also says “This feature isn’t available to users under 18”. ↑c“Only upload photos you have the right to use”. ↑d
  3. 3Google, Generative AI Prohibited Use Policy, last modified 17 Dec 2024, read at source 1 Oct 2026“Impersonating an individual (living or dead) without explicit disclosure, in order to deceive.” ↑
  4. 4Runway, Gen-4 Video Prompting Guide, read in a browser 17 Sep 2026 (as cited on AI images and video for beginners): “Begin with a foundational prompt that captures only the most essential motion to the scene.”↑

Runway, Gen-4 Video Prompting Guide, read in a browser 17 Sep 2026 (as cited on this site’s AI images and video for beginners page).

The comparisons on this page (the film crew with one photo, the director shouting “action”) are this site’s own ways of explaining, not claims from these sources.

Free printable AI cheat sheets, and AI news every week.

Sixteen one-page AI guides and checklists in one free PDF, no sign-up. And if you want to keep up, the newsletter covers what changed in AI each week, in plain English, with the sources.

SUBSCRIBE FREE ↗ GET THE CHEAT SHEET PACK
aipromptgeneer://$ about this siteAI PROMPT GENEERAI explained in plain English. Every claim links to its source, with the date it was checked.NEWSLETTERWhat changed in AI each week, in plain English.SUBSCRIBE FREE ↗AI PROMPT GENEER · MACHINE ASSISTANCE DISCLOSEDEVERY ARTICLE CARRIES ITS OWN CHECKED DATE