How do I make a product video from a photo?

Image-to-video AI takes one still photo and generates a short clip of motion around it. The model invents plausible movement, such as a slow rotation, a camera push-in or liquid pouring, instead of filming anything new. The input is one sharp photo: the same photo you would use on the product page.

Each run makes one short clip. For a longer video, generate several clips from different prompts and join them in a video editor.

1. Start from a sharp, well-lit photo: the whole product in frame, in focus, on a plain or simple background. If the background is busy, remove it first (3 credits on the free plan, free on every paid plan). If you only have a rough phone shot, make a clean studio or lifestyle version with product photos, from 2.5 credits an image.
2. Upload it to an image-to-video tool. In TikTok's Symphony Creative Studio, the Image to Video tool takes an image with your product, a model or a scene.
3. Describe the motion: a slow 360° turn, a gentle zoom, a product being poured or opened. Describe the movement, not the product.
4. Pick the length and, where the tool offers it, the shape: vertical 9:16 for TikTok and Reels. Symphony lets you choose the AI model and the video duration.
5. Generate two or three takes and keep the best, since motion models vary from run to run. Symphony takes a few minutes per clip and can sync the result to TikTok Ads Manager.

Image-to-video tools charge by plan, by credits or per clip, and their prices change often, so check the tool's own pricing page; TikTok's help page for Symphony's Image to Video does not list a price. Filming a product video adds a videographer, equipment and editing time. An image-to-video tool skips those, because it starts from a photo you already have.

The photo is the part you control. On ProduktShot a standard product image costs 3 credits, and new accounts get 16 credits without a card, enough to try a few source photos before you animate one.

Image-to-video AI does best with simple, continuous motion: a bottle or jar turning to show every side, a garment turning to show its drape, a candle flame flickering, liquid pouring into a glass, or packaging opening partway.

The model invents plausible movement; it does not capture how your product really works. Film a real demonstration when the mechanics matter: assembling a multi-part product, a zip or clasp in use, an exact measurement, or anything the viewer must see behave as it would in hand. A UGC-style video with a person using the product is the other option.

Tools also limit their inputs. TikTok's help page says one of the models in Symphony does not accept images or videos with real faces as references, so a photo of a person wearing the product may not work there.

Short enough to show one clear motion. No platform sets an ideal length for a product clip, but TikTok's creative guidance asks for the offer in the first 3 seconds and the hook in the first 6, so put the product on screen from the first frame. A short clip also loops cleanly on a product page or in a feed. Keep longer formats for an ad that needs time for a hook, a demonstration and a call to action.

On TikTok, yes, when the video is AI-generated or changed by AI beyond minor edits. TikTok's ad policy asks for its AIGC label or a clear disclaimer, caption, watermark or sticker, and rejects or restricts ads with undisclosed AI content; Symphony adds an AI-generated label to every exported video. Meta marks ad images made or significantly edited with AI tools with an "AI info" label. Check each platform's policy before you publish, and see TikTok's ad video specs for the file itself.