Kling Motion
Animate a still character image with the motion of any reference video, keeping the face and expressions intact.
Créé avec Kling Motion
Exemples de rendus. Ouvrez Studio pour générer les vôtres.
Là où Kling Motion excelle
Character Animation
Animate a single character image so it moves exactly like the person in your reference video.
Motion Transfer
Reuse a recorded performance by transferring its movement onto a different character image.
Performance Reuse
Drive a still portrait with reference footage while keeping the original face and expressions consistent.
Points forts
- Transfers movement from a reference video onto a still character image while preserving facial identity and expressions
- Two generations to choose from: v2.6 (cheaper, mature) and v3.0 (newer, sharper)
- Two quality tiers, Std at 720p and Pro at 1080p
- Character orientation can follow either the input image or the reference video
- On v3.0 you can pick the output background source, the reference video or the character image
- Optional text prompt for extra scene direction
Compromis
- Needs both a character image and a motion-reference video; it cannot generate motion from a text prompt alone
- Output length is driven by the reference video, not a duration you select
- Priced per second of output, so longer reference clips cost more
- Background-source control is only available on v3.0
- Resolution tops out at Pro (1080p)
En un coup d'œil
À propos de Kling Motion
Kling Motion is Kuaishou's motion-transfer model for video. Give it one character image and one motion-reference video, and it animates your still character using the movement from the reference clip while preserving the character's facial identity and expressions. It is built for bringing a single image to life with real, recorded motion rather than guessing motion from text.
You can run two generations on HexGen. Kling v2.6 is the cheaper, mature option that introduced motion control, while v3.0 is the newer generation that kie.ai describes as improving facial-identity consistency and motion stability. Each generation offers two quality tiers, Std at 720p and Pro at 1080p, and you can set the character to follow either the input image orientation or the reference video orientation. On v3.0 you can also choose whether the output background comes from the reference video or the character image.
Output is delivered as MP4. Length follows the uploaded reference video rather than a duration you pick, so the clip you provide drives the final result. An optional text prompt is available for extra scene direction, but motion always comes from the reference video, never from the prompt alone.
Points de départ
Copiez, ajustez et lancez. Un bon prompt fait l'essentiel du travail.
Use my character portrait and the dance reference video to make her perform the same routine, keeping her face and expression identical.
Animate this standing character with the motion from the reference clip, character orientation following the reference video.
Transfer the walking motion from the uploaded video onto my character image, output background taken from the reference video.
Payez seulement ce que vous générez. 1 USD = 1,000 HGcoins. Les HGcoins n'expirent jamais et les échecs sont remboursés automatiquement.
Kling Motion face aux autres modèles
Kling Motion is the right pick when you already have a character image and a reference clip and want the recorded motion transferred onto your character. Other Kling video models generate motion from prompts or images instead.
| Modèle | Qualité | Vitesse | Coût | À choisir quand |
|---|---|---|---|---|
Kling Motion Celui-ci Kuaishou | Excellent | Rapide | Coût moyen | Choose this when you want to animate a still character image with the exact motion of a reference video while keeping the face and expressions intact. |
Kuaishou | Excellent | Rapide | Coût élevé | Pick this for general Kling video generation when you do not have a motion-reference clip to drive the character. |
Kuaishou | Très bon | Rapide | Coût moyen | A more affordable Kling video option when you want generation without supplying a reference performance. |
Modèles similaires
Questions fréquentes
It transfers movement from a reference video onto a still character image, animating your character with that motion while preserving the character's facial identity and expressions.