Wan 3 Görselden Video

Sets your uploaded frame in motion as the first frame; duration, resolution, and audio are each chosen separately.

modellane/wan-3-i2v · Image to video · Alibaba

A video model that sets an uploaded frame in motion. The image becomes the first frame and the prompt describes what happens in it: how the camera moves, where the elements go, and how the scene develops. The source image's composition, colours, and characters are preserved, so with a ready image in hand the result is far more predictable than generating from scratch. Duration and resolution are chosen from the parameter list in the run form; audio is optional.

Vendor

Alibaba

Task

Image to video

Output

Video

Price

$0.06 – $0.12 · per second

Longest run

10 seconds

Input parameters

• Source image image — required The frame the generation starts from.
• Prompt prompt — required, at most 20000 characters Describe what you want produced as concretely as you can.
• Duration duration — optional, default: 5, range: 2–10 sn, affects price A longer duration raises the price directly.
• Resolution resolution — optional, default: 480p, options: 480p, 720p, affects price Higher resolution raises the per-second price.
• Generate audio audio — optional, advanced, default: true When off, the video is produced silent.

Recommended use cases

• Animating a still image — Turn a product photo, illustration, or poster you already have into short motion; the composition stays exactly as it is.
• Video version of a campaign image — Derive a short clip for the feed from an approved campaign frame without disturbing its framing.
• Character and brand consistency — Carry over the elements that must stay recognisable, such as a face, a garment, or packaging; text-to-image changes them every time.
• Camera movement tests — Try different camera moves on the same frame and see which one suits the scene.

Where it is strong

• Preserves the first frame's composition, colour, and characters
• Reads the prompt as a description of motion; it does not rebuild the scene
• Gives a far more predictable result than text-to-video
• Can generate a soundtrack that matches the picture

Limits

• A starting frame is required; it does not run on the prompt alone
• The reference image must be JPEG, PNG, or WebP with a long edge between 240 and 8000 pixels; the file size ceiling is written in the field's own hint
• The aspect ratio is not chosen separately; the framing follows the ratio of the uploaded frame
• Does not reliably introduce an object that is not in the source frame

Example prompt

The camera slowly pushes in while the fabric ripples in a soft breeze and dust motes drift through the light

Example request

POST /v1/queue/modellane/wan-3-i2v

Playground

The identifier, price, parameters, and example requests on this page can be read without signing in. Running the model from the panel sends a real generation request and charges the project balance; the playground therefore requires an account.