How to Turn a Product Photo Into a Sales Video With AI

Updated November 1, 2026

A still product photo can become a short sales video without a reshoot. Upload it to Oranzy’s Image-to-Video tool, describe a small movement, and check that the finished clip preserves the product’s shape and branding.

Prepare a product photo

Choose a sharp, well-lit image with the entire product visible. Use a clean background and leave space around the object so camera movement does not crop it. Check that the logo, label, color, and packaging match the item you sell: the video starts from this image, so flaws in the source can carry through.

Decide where the video will appear. Use 9:16 for vertical placements and 16:9 for widescreen placements. If the source photo has a different shape, prepare a version with enough background around the product for the chosen frame.

Create the video in four steps

1. Open Oranzy, select Image-to-Video, and upload your product photo as the starting frame. If you need an account, sign up for 100 bonus tokens.
2. Choose a model, such as Kling, Veo 3, or Seedance. Veo 3 supports sound; check the available settings for your selected model before generating.
3. Set the aspect ratio to 9:16 or 16:9 and choose a duration in the 5–10 second range. Start with 5 seconds when testing a new image or prompt.
4. Enter a motion prompt and generate the clip. Review the result for product shape, label, logo, color, and unwanted changes before using it in an ad or product listing.

Write a controlled motion prompt

Describe the product, one camera movement, and the details that must stay consistent. For example:

A skincare bottle stands upright on a pale stone surface. The camera slowly pushes in while soft light moves across the background. Keep the bottle shape, cap, label, logo, and colors consistent with the source image. Do not add text or extra objects.

Replace the product and setting with your own. For a shoe, try a slow camera move around the stationary shoe; for a packaged product, try a gentle push-in. If the model changes the label, simplify the camera movement and try again. Treat prompt instructions as guidance, not a guarantee that small text or logos will render accurately.

Avoid mistakes and wasted generations

• Using a crowded or blurry source: Start with a clearer photo rather than asking the model to repair important product details.
• Requesting too much movement: A full spin, changing background, and moving props can compete for attention. Test one subtle movement first.
• Ignoring the frame: Preview the 9:16 or 16:9 composition before generating. Keep the product away from the edges to reduce the risk of cropping.
• Trusting the logo without checking: Pause on several frames and compare the label and silhouette with the original photo. Reject clips that distort branding or show a misleading product.
• Changing everything between tests: Adjust one variable at a time, such as the camera movement or model, so you can identify what improved the result.

Finish the clip for your placement

Review the full 5–10 seconds at the size viewers will see it. If you need a price, offer, or call to action, add it in your video editor rather than relying on generated text. Leave room for platform interface elements, and check that any sound suits the placement. Export a version for each required aspect ratio and watch the final file once before publishing.