Kling V3 Motion Control on AI Compare Hub

Kling V3 Motion Control by KwaiVGI transfers choreography, gestures, and motion sequences from one video onto a character you define with a single reference image. Output is 720p (Standard) or 1080p (Pro), with up to 10 seconds in image orientation or 30 seconds in video orientation. Ideal for dance covers, motion-driven character animation, and re-staging existing performances with new subjects.

What you can create

Why creators choose Kling V3 Motion Control

How to generate your first motion-transfer video

  1. Pick your reference motion video. Upload the source clip whose motion you want to transfer. Clear, well-lit motion with a single subject works best. The reference video's length and your character_orientation setting determine the maximum output duration.
  2. Provide your character image. Upload a single full-body or portrait image of the character that should perform the motion. The character's identity, outfit, and proportions are preserved through the output.
  3. Configure mode, orientation, and prompt. Pick Standard (720p) or Pro (1080p), choose image orientation (≤10s) or video orientation (≤30s), and toggle keep-original-sound. Optionally add a prompt to direct background, camera, or stylistic details. Generate and review.

Common questions

What is Kling V3 Motion Control?

Kling V3 Motion Control is KwaiVGI's video-to-video motion-transfer model, available via the Replicate API. It takes a reference motion video plus a single character image and produces a new clip in which the character performs the reference motion. Output is 720p (Standard) or 1080p (Pro), up to 10s (image orientation) or 30s (video orientation).

What inputs are required?

A character reference image and a reference motion video are both required. An optional text prompt can guide background, camera, and stylistic details that aren't dictated by the reference video itself.

How is the output duration determined?

Output duration is bounded by the reference video length and the selected character_orientation. image orientation caps at 10 seconds; video orientation caps at 30 seconds. The credit quote uses the reference video's actual duration when available.

How much does Kling V3 Motion Control cost in credits?

Standard mode (720p) costs 35 credits per second. Pro mode (1080p) costs 54 credits per second. A 10-second clip in Standard is ~350 credits; a 10-second clip in Pro is ~540 credits.

What's the difference between image and video orientation?

image orientation produces portrait-style output focused on the character (max 10 seconds). video orientation follows the reference video's full framing for longer sequences (max 30 seconds). Pick image orientation for short character-focused clips and video orientation for longer scene-style outputs.

Can I keep the audio from the reference video?

Yes — enable "keep original sound" to carry audio from the reference motion video into the output. This is especially useful for dance covers, music-driven motion, and any case where the original soundtrack should stay in sync with the transferred motion.

How can I use Kling V3 Motion Control on AI Compare Hub?

Open the AI Video Generator's Video-to-Video panel and pick "Kling V3 Motion Control" from the model picker. Upload your reference motion video and a single character image, choose Standard or Pro, set orientation, optionally add a prompt and toggle keep-original-sound, then generate.

Key Parameters

For the Use of This Model

Kling V3 Motion Control

Apply the motion of a reference video to a custom character image, producing a new video where the character performs the original motion. Supports realistic humans and illustrated characters; full-body or half-body framing.

Inputs

  • image - Character reference image (drives the visual appearance).
  • video - Reference motion video (.mp4 or .mov).
  • prompt - Optional text guiding background, camera, or motion effects.
  • mode - std (720p) or pro (1080p).
  • character_orientation - image (max 10s, character faces same direction as portrait) or video (max 30s, character follows the reference video).
  • keep_original_sound - Preserve audio from reference motion video.

Try Kling V3 Motion Control on AI Compare Hub

Generate with Kling V3 Motion Control directly on AI Compare Hub. Compare results side by side with other leading AI models using the same prompt.