4 motion-control models that need a still and a driving clip in 2026

Four published rows carry motion-control and requires_image: DreamActor V2 (5 cr), Wan 2.2 14B Move (8 cr, 720p), Kling Video V3 Standard Motion Control (13 cr, 720p), and Kling Video V3 Pro Motion Control (17 cr, 1080p) — a paragraph is not a driving clip.

This is the install list. Each item is a live catalog row whose category is motion-control and whose image gate is on. Motion control needs a driving clip, not a paragraph is the two-file rule; this page is four rows and the one job each is for.

Two files, or it is not this family

Motion control copies timing you already captured and puts it on a subject you supply. You need:

1. A character still — framed like the clip. Full body if the motion uses a full body. A tight headshot against a full-body drive is the most common waste.
2. A driving clip — the walk, the dance, the gesture, the turn. Subject in frame. You have the rights.

Transfer a dance or motion onto my photo is the named agent job for this pair: photo in, driver in, retargeted motion out. AI motion transfer is the tool-shaped door.

Missing the driver is ordinary image-to-video. Missing the still is a generated performer. A sentence about choreography is neither file. The image-to-video tool is the door when there is no driving clip. Mixing those doors is how you pay transfer prices and receive an I2V take.

The four published rows

Same family. Four bills. Do not “upgrade in the prompt.” Change the row.

1. DreamActor V2 — 5 credits, face reenactment

DreamActor V2 is ByteDance face reenactment: transfers facial expressions and head movements from a driving video to a reference image. Category motion-control. requires_image true. Catalog list 5 credits. audio unset. No duration enum. No max-output chip on the record.

This is the face-only row. Headshot still plus a driving clip of a face. A full-body dance on this SKU is the wrong region. Guidance text is not choreography. If the mouth is a later lipsync, that is a different family after you have a plate.

2. Wan 2.2 14B Move — 8 credits, 720p

Wan 2.2 14B Move is Wan motion control with 14B parameters. Category motion-control. requires_image true. Max output 720p. Catalog list 8 credits. audio unset. No duration enum.

Use it when the body has to move and 720p is the ceiling you will ship. Do not brief a client on 1080p from this SKU. Length follows the driving clip. Trim the driver to the move you actually need.

3. Kling Video V3 Standard Motion Control — 13 credits, 720p, silent

Kling Video V3 Standard Motion Control transfers motion from a driving video to a reference image. Category motion-control. requires_image true. Max output 720p. Catalog list 13 credits. audio false. No duration enum. Quality listed: 720p only.

Standard is the resolution flag, not a different dance. The file is silent. Quoted music in the prompt does not become a soundtrack. Character orientation and background source are switches: whose facing wins, whose backdrop wins. Set them. A headshot still plus a full-body driver is the classic waste at 13 credits.

4. Kling Video V3 Pro Motion Control — 17 credits, 1080p, silent

Kling Video V3 Pro Motion Control is the higher-tier sibling: transfer at pro-tier quality. Category motion-control. requires_image true. Max output 1080p. Catalog list 17 credits. audio false. No duration enum. Quality listed: 1080p only.

1080p only. There is no 720p option and no 4K option on this record. Audio is false. This row will not write a score and will not lip-sync a line. If the mouth is the shot, run a lipsync model after you have a plate. Do not prompt Standard into Pro. Change the row.

A paragraph is not a driving clip

The sentence people try: “A woman dances like the reference, energetic, matching the beat.”

There is no reference. There is a hope. The model will generate a dance. It will not generate the dance you saw on TikTok or the one in an unlinked Drive folder. These four rows are the transfer family. Bring the clip.

An optional prompt can name wardrobe, backdrop, mood. It does not replace the driving video. Writing “subject leaps” while uploading a clip of someone standing still is how you spend the listed credits on a standing still.

A clip you do not have rights to is not a driver. A private person’s face as the still, without a replica grant, is a likeness problem before it is a generate. Minors are a hard stop.

How to pick without mixing I2V

You have	You need	Row
Headshot still + face driver	Expression and head motion, 5 credits	DreamActor V2
Full-body still + body driver, 720p is enough	Transfer at 8 credits	Wan 2.2 14B Move
Full-body still + body driver, silent 720p	Kling transfer, Standard	Kling V3 Standard Motion Control
Full-body still + body driver, silent 1080p	Kling transfer, Pro	Kling V3 Pro Motion Control
A still you love and no clip	Invented camera, not a copy	Image-to-video, not this list
A paragraph about a dance	Nothing yet	Film or find the clip, or drop the job

Two files — a character still plus a motion clip you would actually copy — and this is the job. One still, and you are in image-to-video. If the movement is “slow push-in, soft smile,” do not buy a motion-control row.

FAQ

Can I describe the dance instead of attaching a reference clip?

Not for these four. The driver is the motion. A paragraph about choreography is a different generate, and a worse one when timing is the point. Transfer a dance or motion onto my photo expects both files.

Why is Standard 13 credits and Pro 17 if they do the same transfer?

On this catalog, Standard is 720p at 13 credits, audio off. Pro is 1080p at 17 credits, audio off. Resolution and sticker, not a different dance. Do not prompt Standard into 1080p.

Does DreamActor V2 output a full-body dance at 5 credits?

No. The listed job is face reenactment: expressions and head movements onto a reference image. Full-body copy belongs on Wan Move or a Kling V3 motion-control row.

Is this the same as uploading one image to image-to-video?

No. One image is open-ended motion. Two files is a bounded copy. Image-to-video is the door when you have no driver. These four rows are the door when you do.