Guides · Generative AI

How to Make an AI Video from a Photo: Free and Paid Ways, Step by Step

By Ravindra Bagale, trainer · LinkedIn · Published: 19 October 2026 · Last updated: 19 October 2026 · 9 min read

To turn a photo into an AI video, upload a clear photo to an image-to-video tool and describe the motion, the camera move and the sound you want. The easiest free way is Google Photos: Create → Photo to video, which makes short six-second clips with a daily limit. For longer, more controlled videos, the Gemini app uses Google's Gemini Omni video model, which needs a Google AI plan (October 2026). OpenAI's Sora app has been shut down.

1. Suppose we sell sarees on Instagram

Suppose we are Kavita. We sell Paithani sarees from our shop in Yeola, near Nashik, and take orders on Instagram. Our photos get likes, but reels get far more reach. We have no camera crew and no time to shoot a video.

We have one good photo: a green Paithani saree spread on a wooden table, with the gold border clearly visible.

We try a weak prompt in a video tool:

make video

What happens: the tool adds random movement. The saree may melt, the border pattern may change, or a hand may appear from nowhere.

Then we write a clear prompt:

The saree stays still on the table. A soft breeze slowly lifts one corner of the
pallu so the gold border shines in the light. The camera slowly moves closer.
Soft sitar music in the background. Calm, rich mood.

Expected result (yours will be similar, not the same): a short clip where only the pallu corner moves and the camera glides in. The saree design stays the same, because we told the tool exactly what moves and what does not.

2. What is image-to-video?

An image-to-video tool takes one photo as the first frame and imagines what happens next. It does not record anything. It draws each new frame, guessing how things would move.

This is why:

• Small, simple movements look real. Hair in the wind, a smile, clouds moving, steam from chai.
• Big, complex movements often break. People dancing, hands doing work, text on signboards and fast camera spins.
• Every try is different. Make 2 or 3 versions and keep the best one.

3. Which tool should I use? (October 2026)

Two notes before you choose:

• OpenAI's Sora app is gone. OpenAI shut down the Sora web and app experience on 26 April 2026, and the Sora API ended on 24 September 2026. Old tutorials that tell you to use Sora no longer work.
• Google's own pages say these features are 18+, need a personal Google account, and are not in every country yet. If you do not see the option, it is not in your account yet.

4. Method 1: a free clip in Google Photos (step by step)

1. Open the Google Photos app on your phone (Google's help page shows these steps for Android).
2. Make sure the photo is backed up (Photos needs this for AI features).
3. Tap Create at the bottom, then Photo to video.
4. Tap Try now if you see it, then Add photo and choose your photo.
5. Choose an effect: Subtle movements (a small natural move, like a smile) or I'm feeling lucky (bigger moves or effects like confetti).
6. Or type your own short prompt (this option is for users 18 and over).
7. Wait a few moments. Then tap Save to keep the video or Share to post it.

Note: a landscape photo is cropped to a vertical video, so keep the main subject in the centre of your photo.

5. Method 2: a longer clip in the Gemini app

1. Open the Gemini app or gemini.google.com and sign in with an account that has a Google AI plan.
2. Choose the video option from the tools menu (the names in the menu can change; look for video creation).
3. Upload your photo (Gemini Omni can use up to five photos).
4. Type your prompt using the formula in section 6.
5. Wait for the clip. Video takes longer than an image.
6. Now edit it by chatting: "Keep everything, but make the camera move slower", or "What happens next? Extend the scene by a few seconds".
7. Download the clip and add your own text or music in an editor if needed.

6. The 4-part video prompt formula

Copy this template:

Subject: [what is in the photo] stays [still / the same].
Motion: [one simple movement], [slowly / gently].
Camera: [slow push-in / slow pan / still].
Sound and mood: [music or sounds], [mood].

Keep it to one main movement. One clear move looks real. Five moves in one clip usually look strange.

7. Worked examples: three photos, three videos

Example 1, a chai stall photo (Google Photos, free). We choose "Subtle movements" on a photo of a glass of cutting chai on a stall counter in Pune.

Expected result (similar, not the same): steam rises from the chai and light flickers slightly. Six seconds, perfect as a story.

Example 2, a travel photo of Sinhagad fort (text prompt).

The fort walls stay the same. Clouds drift slowly across the sky from left to right
and the grass moves gently in the wind. Camera stays still. Sound of wind.
Peaceful monsoon morning mood.

Expected result: the sky and grass move; the fort does not wobble.

Example 3, a product photo of a steel water bottle (Gemini).

The bottle stays exactly the same, including the logo. The bottle slowly turns
around once on the table. Camera stays still. Clean studio light, soft electronic
music, modern mood.

Then a follow-up in the same chat: "Keep everything, but add drops of water sliding down the bottle."

Check before posting: pause the video at the start, middle and end. Look at the logo, text and edges. If the logo changes shape, ask again or use a shorter, simpler movement.

8. 8 copy-paste motion prompts

1. "The person smiles gently and blinks once. Camera stays still. Soft background music."
2. "Hair moves slightly in a breeze. Everything else stays still. Camera slowly moves closer."
3. "Flames of the diyas flicker softly. Camera slowly pans from left to right. Festive mood."
4. "Rain falls outside the window; the room stays still. Sound of rain. Calm mood."
5. "Waves roll gently onto the beach. Camera stays still. Sound of waves."
6. "Leaves fall slowly from the tree. Camera slowly rises upward. Autumn mood."
7. "The shop's lights turn on one by one as evening falls. Camera stays still. Street sounds."
8. "Steam rises from the plate of hot pav bhaji. Camera slowly moves closer. Busy restaurant sounds."

9. Safety, consent and labels (read this before you post)

• Only animate photos you own or have permission to use. Animating a stranger, a classmate or a celebrity can hurt people and can be illegal.
• Be careful with photos of people who have passed away. Ask the family first. Many people find it painful.
• Never make a person "say" something they did not say. That is a deepfake, and it is how many scams work. See our guide on deepfake and voice clone scams.
• Label AI videos. Google marks its AI video with an invisible SynthID watermark, and you can ask the Gemini app whether a file was made with Google AI. India's IT Rules amendment of 2026 asks platforms to clearly label AI-made video. On YouTube, turn on the "altered or synthetic content" disclosure when a video looks real.
• Do not use AI clips as proof of anything, like an accident, a delivery, or a product defect.

10. Beginner mistakes and quick fixes

11. Try at home

Recap

Make one six-second clip today from a photo of your city or your favourite food. Small, simple motion is where every good AI video starts.

Make the photo first: Gemini Nano Banana Photo Editing Prompts

Ravindra Bagale, trainer: linkedin.com/in/ravindra-bagale. Kavita, her shop and the products are made up for learning. Tool names, plans, limits and the Sora shutdown dates were checked on Google and OpenAI help pages on 19 October 2026; they change often. AI videos change on every try, so your result will be similar, not the same.

Related: Image-to-Video Explained: Direct an AI Clip from One Photo