Image-to-video takes a still as frame one and a prompt as stage direction. The still is the contract: who is in the shot, what the product looks like, where the camera already sits, what the light is. The prompt should say what moves. Re-describing the image is how the model tries to satisfy a sentence that is already true and starts walking the frame toward the wording. A pack shot is the commercial version of this contract — the SKU must still be the SKU at second four.
A reference image is a different slot. It is identity without locking the frame; the model may re-angle, re-light, re-size. First-last-frame is a contract at both ends. Motion control's driving clip is a contract on pose, not on the still. Mixing those jobs — a pack still in a reference slot, or a prompt that asks for a new label — is how the logo crawls. Seed-lock the take once the still is right, then change one clause of motion.
The still has to be good enough to obey. Soft type, a cropped box, a filtered lifestyle crop: the clip will inherit that. Retry the still before you retry the video. If the generate breaks the contract, regenerate the shot. Covering a melted label with a caption is not a hold.
In practice
- Write motion, not a second description of the still. The picture already decided the product and the face.
- Approve readable type on the still before you animate. The clip will not sharpen a soft label.
- If identity or packaging drifts, treat it as a broken contract and re-roll — do not grade it away.
Image-to-video models
Catalog entries that animate a still you supply. 67 of the 351 models in the Versely catalog qualify.
| Model | Provider | Type |
|---|---|---|
| Happy Horse 1.1 Image to Video | Alibaba | Video |
| Vidu Q3 Image to Video | Vidu | Video |
| Pixverse 5.6 Image to Video | Pixverse | Video |
| Kling O3 Pro Image to Video | Kling | Video |
| Cosmos 3 Super Image to Video | NVIDIA | Video |
| Kling O3 Standard Image to Video | Kling | Video |
| Wan 2.7 Image to Video | Wan | Video |
| Kling Video V3 Standard Image to Video | Kling | Video |
Browse all 29 spec pages for full settings, resolutions and credit costs.
The mistake to avoid
Prompting a new camera, a new pack, and a new face on an I2V job and calling the mismatch a model bug. You rewrote the contract in the prompt.
Where you will run into it
- AI Video Generator — Text-to-video, image-to-video, and story chaining in one studio.
- AI UGC Video Generator — UGC ads at the speed and price of a prompt.
- Kling O3 Pro Image to Video — Kling O3 Pro image-to-video with advanced reasoning-enhanced generation and camera controls
Related terms
Image-to-video
Image-to-video animates a still you supply: the picture becomes the opening frame, and the prompt describes only what happens next.
Reference image
Reference image meaning: a picture supplied with the prompt so the model can copy identity, product, or style without using it as a frame.
Pack shot
A pack shot is the product still that has to stay readable — label, shape, colour — the plate image-to-video is not allowed to melt.
Text-to-video
Text-to-video is generation from a written prompt alone — you describe a shot, the model invents every frame of it, and no image or footage goes in.
Text-to-image
Text-to-image renders a still picture from a written description, with no picture going in.
The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync, in your browser or on your phone.
Free on the web and on iPhone, one account.