Kwaivgi Kling Elements API Documentation

Kwaivgi Kling Elements API Documentation

Playground

Kling Elements creates custom AI elements from reference images for video generation. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Kling Elements creates custom AI elements from reference images for video generation. Define reusable visual elements — characters, clothing, objects, or styles — that can be consistently applied across multiple Kling video generations.

Why Choose This?

• Reusable AI elements Create custom elements once, use them across multiple video generations for consistent results.

• Multi-reference support Upload multiple reference images to better capture the element’s visual characteristics.

• Character consistency Maintain the same character appearance, clothing, or style across different videos.

• Simple workflow Define elements with a name, description, and reference images — ready to use in video generation.

• Affordable pricing Just $0.01 per element creation.

Parameters

How to Use

1. Enter element name — give your element a clear, descriptive name.
2. Write description — describe the element’s visual characteristics in detail.
3. Upload primary image — provide the main reference image.
4. Add reference images — upload additional images showing the element from different angles or contexts.
5. Add tags (optional) — categorize your element for easier management.
6. Run — submit to create your custom AI element.
7. Save the element_id — use this ID in Kling video generation to apply the element.

Pricing

Best Use Cases

• Character Consistency — Create character elements for consistent appearance across video series.
• Wardrobe Management — Define clothing elements for virtual try-on or fashion videos.
• Brand Assets — Create reusable brand elements (logos, mascots, products) for marketing videos.
• Style Templates — Define visual styles that can be applied consistently.
• Product Visualization — Create product elements for e-commerce video content.

Pro Tips

• Use multiple reference images showing the element from different angles for better recognition.
• Write detailed descriptions including colors, textures, and distinctive features.
• Keep element names clear and descriptive for easy identification later.
• Use tags to organize elements by project, category, or use case.
• Save the returned element_id — you’ll need it to use the element in video generation.

Notes

• All required fields (name, description, image, element_refer_list) must be provided.
• The output includes an element_id that can be used in Kling video models.
• Ensure uploaded image URLs are publicly accessible.
• Multiple reference images improve element consistency.

Related Models

• Kling Video O3 Pro Image-to-Video — Use created elements in video generation.
• Kling Video O3 Pro Reference-to-Video — Reference-based video generation.
• Kling V3.0 Pro Image-to-Video — High-quality image-to-video generation.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "name": "Example",
  "description": "A clear example input",
  "image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
  "element_refer_list": [
    "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg"
  ]
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/kwaivgi/kling-elements" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout|deleted) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    *) sleep 2 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters