Kling 3.0 Motion Control API
Price: $0.015(~ 1 credits) per request
Highest stability with guaranteed 99.9% uptime. Recommended for production environments.
Use the same API endpoint for all versions. Only the model parameter differs.
Choose whether to create element from images or video
Upload a video showing the subject clearly — must contain a visible human face (max 8s, 720-2160px)
Click to upload or drag and drop
Supported formats: MP4, MOV, WEBM
Maximum file size: 100MB; Maximum files: 1
Name of the subject element (max 20 characters)
Description of the subject element (max 100 characters)
No video URL found in response
Check console for details
History
Max 20 items0 running · 0 completed
Billing Rules
- •Fixed price: 1 Credit per element creation
- •Full refund on failure
- •Estimated processing time: ~10 minutes
Pricing
| Model | Mode | Type | Price |
|---|---|---|---|
| Kling Custom Element | Element Creation | Element Creation | $0.015/ request(1 Credits) |
If it's down, we automatically use the next cheapest available—ensuring 99.9% uptime at the best possible price.
Transfer reference-video motion to a character image through EvoLink's unified API, with orientation control, optional original audio, and async delivery for production workflows.
- Starts at $0.1134/s
- 720p / 1080p
- Inputs: image + video
- Up to 10s / 30s

Reference video motion, image character guidance, and async delivery in one production-ready route.
Explore related Kling APIs
Kling 3.0 API
Use the main Kling 3.0 route for text-to-video and image-to-video generation through EvoLink's unified API.
Kling O3 API
Use Kling O3 when you need reference-to-video, video editing, or one route that covers four video workflows.
Kling 3.0 Motion Control API use cases and controls
Kling Character Animation API
Animate a still character image from a reference performance video without manual rigging or frame-by-frame editing. This workflow is useful for creator tools, virtual avatars, short-form content, and character-driven video production where natural human motion matters.

Kling Dance and Gesture Transfer API
Transfer dance moves, gestures, and body performance from one clip to a different character for social media production, marketing assets, and creator workflows. Teams can turn a single motion source into multiple character-led outputs for faster iteration across campaigns.

Kling Motion Control API with Orientation Control
Control whether the generated character follows the direction of the source image or the reference video, then choose 720p or 1080p output based on budget and production needs. This gives developers a more structured motion-control pipeline than generic image-to-video generation.

Why teams use Kling 3.0 Motion Control through EvoLink
Kling 3.0 Motion Control gives teams a more controllable way to turn real performance footage into character animation, with predictable per-second pricing and a production-ready API workflow.
More controllable than generic video generation
Instead of asking a model to invent motion from scratch, Motion Control uses a reference performance video as the motion source. That makes output behavior easier to steer for character-driven workflows.
Faster production without manual rigging
Teams can turn a still image plus a motion clip into usable animated output without building a traditional animation pipeline. This reduces setup friction for prototypes, campaign assets, and creator tools.
Predictable API pricing for repeated workflows
Per-second billing, 720p and 1080p modes, and explicit duration limits make it easier to estimate cost before production. That matters for batch jobs, user-generated workflows, and automated content pipelines.
How to integrate Kling 3.0 Motion Control API
From reference inputs to delivered animation in three practical steps.
Prepare a clean character image and motion source video
Upload a reference image for the target character and a reference video containing the performance you want to transfer. Clear body visibility and stable source motion improve output consistency.
Submit an async generation request with motion settings
Send the image, video, quality mode, and orientation parameters to the API. The task runs asynchronously, which makes it easier to integrate into production pipelines, queue systems, and automated workflows.
Poll task status, review output, and iterate
Check task status until the result is ready, then review the generated clip and refine inputs when needed. This iteration loop works well for creator products, internal tools, and repeatable video workflows.
Kling 3.0 Motion Control API capabilities
Reference-driven character animation with production-ready controls
Reference video motion transfer
Use a source performance video to drive movement in a target character image, making motion output more repeatable and easier to control than prompt-only generation.
Character orientation control
Choose whether the output character follows the facing direction of the reference image or the source video, depending on the type of motion transfer you need.
Custom Element support for character consistency
Combine Motion Control with Kling Custom Element workflows when you need more stable character identity across repeated generations and production variations.
Original sound preservation
Keep the original sound from the reference video when the workflow calls for motion plus source-audio continuity in the final output.
Per-second pricing with explicit output limits
Estimate cost more easily with per-second billing, visible duration constraints, and clear quality tiers. This helps teams plan usage for creator tools and automated pipelines.
720p and 1080p delivery modes
Choose a lower-cost or higher-quality output path based on production goals, turnaround needs, and budget sensitivity.
Kling 3.0 Motion Control API FAQ
Everything you need to know about the product and billing.
API Reference
Select endpoint
Authentication
All APIs require Bearer Token authentication.
Authorization:
Bearer YOUR_API_KEY/v1/videos/generationsCreate Element
Kling Custom Element (kling-custom-element) creates reusable subjects (elements) from reference videos. Created elements can be referenced in Kling V3 Motion Control, Kling O3 series, and Kling V3 Image-to-Video video generation via element_list for consistent character appearance.
Asynchronous processing mode, use the returned task ID to query status.
Upon completion, result_data will contain the element_id for use in video generation.
Important Notes
- This model creates reusable elements (subjects), not videos. No prompt/duration/quality/aspect_ratio parameters needed.
- The reference image or video must contain a clearly visible human face.
- Fixed pricing: 1 Credit per element creation. Full refund on failure.
- Estimated processing time: approximately 10 minutes.
- After successful creation, use the returned element_id in Kling V3 Motion Control, Kling O3 series, and Kling V3 Image-to-Video video generation via element_list.
Request Parameters
modelstringRequiredDefault: kling-custom-elementModel name for custom element creation.
kling-custom-elementmodel_params.element_namestringRequiredName for the element (subject) being created.
Notes
- Max 20 characters
MyCharactermodel_params.element_descriptionstringRequiredDescription of the element, used to help the model understand the subject characteristics.
Notes
- Max 100 characters
A young male character with short hair, wearing a white T-shirtmodel_params.reference_typestringRequiredType of reference material used to create the element.
| Value | Description |
|---|---|
| video_refer | Use reference video to create element |
video_refermodel_params.element_video_listobjectOptionalReference video for element creation. Required when reference_type is video_refer.
Notes
- Conditionally required: must be provided when reference_type = video_refer
- video_url (string): URL of the reference video
- The reference video must contain a clearly visible human face
{"video_url": "https://example.com/reference.mp4"}model_params.element_voice_idstringOptionalVoice ID to assign to the element. The voice will be used when the element speaks in generated videos.
Notes
- Optional parameter — only supported when reference_type = video_refer
- Not available for image_refer (image-based element creation does not support voice assignment)
- See the Voice ID Reference section below for all available voice IDs
829824295735410756callback_urlstringOptionalHTTPS callback address after task completion.
Notes
- Triggered on completion, failure, or cancellation
- HTTPS only, no internal IPs
- Max length: 2048 chars
- Timeout: 10s, Max 3 retries
https://your-domain.com/webhooks/element-task-completed