Fashion video starts from a still of a model wearing a garment and returns a short clip of that model moving. The garment type sets the frame, so a saree is not cropped like a t-shirt. This guide is the sequence the studio actually uses.
What to upload
Open the AI fashion video generator. You need one photo of a model in the garment, a saved AI character, or a still you already generated. Leave headroom and floor in the frame so the model can move without hitting an edge. A sharp, evenly lit source holds up better than a dim or filtered one.
If you only have a flat-lay, make an on-model still first with virtual try-on, then bring that frame here.
How to generate the clip
- Bring the still. Upload it, pick a saved character, or pull one from your gallery.
- Name the garment. Womenswear, menswear, or androgynous. This sets framing: chest to head for tops, head to feet for sarees, lehengas and suits, waist to feet for skirts and trousers. Labelling a floor-length dress as a top will crop it at the waist.
- Set look and motion. Pick a style preset and a motion (walk, turn, pose, spin, idle), use a one-click template, or write a custom prompt.
- Choose output, then generate. Engine, aspect ratio (six options, 9:16 default), resolution (480p, 720p, 1080p), and 5 or 10 seconds. Grok caps at 6 seconds. The studio shows the credit estimate before you commit. Failed generations refund automatically.
Which engine and motion
Seedance holds identity and garment detail and is the default for anything that will sit next to the product. Kling gives smoother camera movement. Grok is the short iteration tool. Match motion to the piece: spin for flare or drape, walk for full length, pose for jewelry and detail, idle for a quiet loop.
Fashion video is not on the free plan. Iterate at 5 seconds and 720p, then re-render keepers at 1080p. Rates live in the studio estimate, not on this page, because they can change.
What this is not
It is not a campaign film and not a fit or sizing tool. The motion is a plausible render. For a talking presenter, use the UGC video generator instead. For any still that is not a worn garment, use image to video.
Frequently asked questions
Why does it ask what clothing is in the photo?
Garment type sets the camera frame. Tops are chest to head. Sarees, lehengas and suits are head to feet. The right type is what stops a floor-length piece being cropped at the knee.
Is fashion video on the free plan?
No. It runs on a paid credit pack. Credits are refunded if a generation fails. The studio shows the estimate next to the engine picker.
How is this different from image to video?
Image to video animates any still. Fashion video takes clothing type as an input, derives framing from it, and offers fashion-specific motions and presets.
What lengths and ratios can I export?
Clips are 5 or 10 seconds, except Grok, which caps at 6. Ratios are 9:16, 3:4, 1:1, 4:3, 16:9 and 21:9. Resolutions are 480p, 720p and 1080p.