Kling Advanced Elements creates custom AI elements from reference images or videos for consistent character and object appearance across Kling video generations. Supports multi-image elements with frontal and reference images, video character elements, and optional voice binding. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Idle
{
"element_id": 307212335131305,
"element_name": "handsome man",
"element_type": "image_refer",
"element_description": "a cool guy"
}$0.01per run·~100 / $1
ExamplesView all
Related Models
README
Kling Advanced Elements
Kling Advanced Elements creates custom AI elements from reference images or videos for consistent character and object appearance across Kling video generations. Define an element with a name, description, and reference material — the model returns a reusable element ID that can be referenced in any Kling generation to maintain identity across clips. Supports both image-based and video-based element creation, with optional voice binding for speaking characters.
Why Choose This?
-
Two reference modes Choose image_refer (frontal image + up to 4 additional reference images) or video_refer (reference video) to best match your source material.
-
Multi-image support Capture different angles, expressions, and styles with a frontal image plus up to 4 additional reference images for accurate character consistency.
-
Video character elements Define a character's full appearance and motion style from a reference video for more dynamic identity capture.
-
Voice binding Optionally attach a voice ID to the element for talking avatar and dialogue-driven video workflows.
-
Reusable across generations Created elements can be referenced by ID in any Kling video generation — use the same character across unlimited clips.
Parameters
| Parameter | Required | Description |
|---|---|---|
| name | Yes | Element name. Max 20 characters. |
| description | Yes | Element description. Max 100 characters. |
| reference_type | Yes | Reference mode: image_refer (default) or video_refer. |
| frontal_image | Yes (if image_refer) | Front-facing reference image. Required when reference_type is image_refer. |
| refer_images | No | Additional reference images (2–4) from different angles or expressions. |
| element_video_list | Yes (if video_refer) | Reference video defining the character's appearance. Required when reference_type is video_refer. |
| voice_id | No | Voice ID to bind to the element for speaking characters. |
| tag_list | No | Custom tags for organizing and categorizing elements. |
How to Use
- Enter a name — give your element a clear, identifiable name (max 20 characters).
- Write a description — describe the character's appearance, style, and key traits (max 100 characters).
- Select reference_type — choose image_refer for image-based creation or video_refer for video-based.
- If image_refer — upload a frontal_image (required) and optionally add 2–4 refer_images from different angles.
- If video_refer — upload one reference video in element_video_list.
- Add voice_id (optional) — attach a voice ID for speaking character workflows.
- Add tag_list (optional) — add custom tags to organize your element library.
- Submit — save the returned element ID for use in Kling video generations.
Pricing
| Reference Type | Cost per Element |
|---|---|
| image_refer | $0.010 |
| video_refer | $0.015 |
Best Use Cases
- Consistent character series — Create a reusable character ID to maintain identity across multiple Kling video generations.
- Fashion & wardrobe elements — Define clothing and styling elements for consistent use in fashion video content.
- Brand assets — Build reusable brand mascots, logos, and product elements for marketing video workflows.
- Talking avatar workflows — Combine element IDs with voice IDs for dialogue-driven character video generation.
- E-commerce product elements — Define product elements for consistent product video content at scale.
Pro Tips
- Use clear, well-lit frontal and profile images for the most accurate character identity capture.
- For video_refer mode, use a short clip that clearly shows the character from multiple angles.
- Give elements descriptive names and tags to keep your library organized as it grows.
- Once an element is created, write its name naturally in your generation prompt and enter the element ID in the element_list field — no special characters required.
Notes
- name, description, and reference_type are always required.
- image_refer mode requires at least a frontal_image; refer_images are optional (2–4 additional images).
- video_refer mode requires exactly 1 reference video and costs 1.5× the image_refer price.
- Voice binding is optional and available for both reference types.
- Voice IDs can be obtained through the voice-related API — see the Voice Guide for details.
Related Models
- Kling Elements — Standard element creation for Kling video models.
- Kling Video O3 Pro Text-to-Video — Use your elements in O3 Pro text-to-video generation.
- Kling Video O3 Pro Image-to-Video — Use your elements in O3 Pro image-to-video generation.
Kling Elements Advanced API — Quick start
Grab a WaveSpeedAI API key, then call POST https://api.wavespeed.ai/api/v3/kwaivgi/kling-elements-advanced with your input as JSON. The endpoint returns a prediction id; poll the prediction endpoint until status flips to completed, then read the output URL from data.outputs[0]. Examples for Kling Elements Advanced below.
HTTP example
# Submit the prediction
curl -X POST "https://api.wavespeed.ai/api/v3/kwaivgi/kling-elements-advanced" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $WAVESPEED_API_KEY" \
-d '{
"reference_type": "image_refer",
"refer_images": [
""
]
}'
# Response includes a prediction id. Poll for the result:
curl -X GET "https://api.wavespeed.ai/api/v3/predictions/{request_id}/result" \
-H "Authorization: Bearer $WAVESPEED_API_KEY"
# When status is "completed", read the output from data.outputs[0].Node.js example
// npm install wavespeed
const WaveSpeed = require('wavespeed');
const client = new WaveSpeed(); // reads WAVESPEED_API_KEY from env
const result = await client.run("kwaivgi/kling-elements-advanced", {
"reference_type": "image_refer",
"refer_images": [
""
]
});
console.log(result.outputs[0]); // → URL of the generated outputPython example
# pip install wavespeed
import wavespeed
output = wavespeed.run(
"kwaivgi/kling-elements-advanced",
{
"reference_type": "image_refer",
"refer_images": [
""
]
}
)
print(output["outputs"][0]) # → URL of the generated outputKling Elements Advanced API — Frequently asked questions
What is the Kling Elements Advanced API?
Kling Elements Advanced is a Kuaishou model for AI inference, exposed as a REST API on WaveSpeedAI. Kling Advanced Elements creates custom AI elements from reference images or videos for consistent character and object appearance across Kling video generations. Supports multi-image elements with frontal and reference images, video character elements, and optional voice binding. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing. You can call it programmatically or try it from the playground above.
How do I call the Kling Elements Advanced API?
POST your input parameters to the model's REST endpoint (shown in the API tab of this playground) with your WaveSpeedAI API key in the Authorization header. Submission returns a prediction ID; poll the prediction endpoint until status flips to "completed", then read the output URL from the result. The playground generates a ready-to-paste code sample in Python, JavaScript, or cURL for whatever inputs you've set. Full request/response shape is documented at https://wavespeed.ai/docs/docs-api/kwaivgi/kwaivgi-kling-elements-advanced.
How much does Kling Elements Advanced cost per run?
Kling Elements Advanced starts at $0.010 per run. That figure is the base price — the final charge scales with the parameters you set in the form (output size, length, count, references, or whatever knobs this model exposes), so a higher-quality or larger output costs more than a minimal one. The exact cost for your current input is shown live next to the Generate button before you submit, and the actual per-call charge is recorded on the prediction afterwards.
What inputs does Kling Elements Advanced accept?
Key inputs: `description`, `element_video_list`, `frontal_image`, `name`, `refer_images`, `reference_type`. The full JSON schema (types, defaults, allowed values) is rendered above the Generate button and mirrored in the API reference at https://wavespeed.ai/docs/docs-api/kwaivgi/kwaivgi-kling-elements-advanced.
How do I get started with the Kling Elements Advanced API?
Sign up for a free WaveSpeedAI account to claim starter credits, copy your API key from /accesskey, then call the endpoint shown in the API tab of the playground. The playground also auto-generates a code sample in Python, JavaScript, or cURL for the parameters you've set.
Can I use Kling Elements Advanced outputs commercially?
Commercial usage rights depend on the model's license, set by its provider (Kuaishou). The license summary appears on the model card above; see WaveSpeedAI's Terms of Service for platform-level conditions.