Home / Guides / Make a product video from a photo

How to make a product video from a photo with AI

6 min read · Published 2026-10-04

You do not need a studio to get a short product video. If you have one good photo, you can generate a few seconds of motion, check it against the real product, and finish it in any editor. This guide walks through the full workflow and the mistakes that cost the most time.

Step 1: Prepare the source photo

The photo decides how faithful the video can be, so spend your effort here. A sharp, evenly lit image with the whole product in frame works best, on a simple background that is not close in color to the product.

Use the highest resolution you have, not a compressed thumbnail.
Keep the product centered with some space around it, so the camera has room to move.
Remove clutter, props and reflections that you do not want animated.
If the product has a label or logo, make sure it is clearly visible and facing the camera.

Step 2: Choose the shot

Pick one shot per clip. Short product videos usually come from a small set of shots, and each one suits a different goal.

Slow push-in: draws attention to the product. A safe default.
Partial orbit (30 to 90 degrees): shows depth and shape.
Light sweep or reflection: highlights materials like glass, metal or ceramic.
Pull-back reveal: starts on a detail and widens, good as an opening.

Step 3: Write the prompt

Use the three-part structure from our prompt guide: what moves, how the camera moves, and what must stay unchanged.

A soft light sweep moves across the bottle while the camera slowly pushes in. Keep the shape, label, logo and colors unchanged. No spinning, no new text, no extra objects.

Step 4: Choose format and duration

Vertical 9:16 for TikTok, Reels, Shorts and most paid social.
Square 1:1 for feeds and marketplace galleries.
Horizontal 16:9 for YouTube and website banners.
Start with a short duration of roughly 5 to 8 seconds. Short clips drift less and cost less to test.

Step 5: Generate a few versions

Do not judge the technique by one result. Generate several versions of the same prompt, then change one thing at a time if you need to improve. Keep the clip that looks most like the real product, not the one with the most dramatic motion.

Step 6: Check accuracy

Compare the first, middle and last frame of each clip against your original photo.

Is the label, logo and color still correct?
Have proportions or small parts such as caps and handles changed?
Does any new text or object appear that was not in the photo?
Does the motion look natural from start to end?

Step 7: Finish in an editor

Add the things AI video does poorly: readable text, prices, brand overlays, music and a call to action. If you need a longer video, join several short clips instead of generating one long one.

Common mistakes

Starting from a low-quality photo and hoping the model will fix it.
Asking for too much motion in one clip.
Letting the model add text to the packaging.
Skipping the frame-by-frame check and publishing a clip with a changed logo.