Kling Video Generation API Promotional Copy

Kling is a video foundation model launched by Kuaishou Kling AI, one of the strongest domestic models. The latest V3 series is the first to achieve “flexible 3–15 second durations + synchronized native sound generation + multi-shot orchestration,” building into the model requirements that previously could only be handled in post-production. Ace Data Cloud has integrated the full Kling series at the earliest opportunity.

Supported Models

Model Duration Features
kling-v3 Any integer from 3–15s Flagship, flexible duration + native sound + multi-shot
kling-v3-omni Any integer from 3–15s V3 multimodal version, supports image and video references
kling-v2-6 5s / 10s The only non-v3 model that supports sound generation
kling-v2-5-turbo 5s / 10s Ultra-fast version
kling-v2-1-master 5s / 10s Previous-generation flagship
kling-v2-master 5s / 10s Classic stable version
kling-v1-6 5s / 10s Compatible with legacy workflows
kling-o1 5s Independent O1 model, supports multi-image references and video editing

Call Examples

# Flexible duration + native sound effects (exclusive to V3)
curl -X POST 'https://api.acedata.cloud/kling/videos' \
  -H 'authorization: Bearer {YOUR_TOKEN}' \
  -H 'content-type: application/json' \
  -d '{
    "model": "kling-v3",
    "action": "text2video",
    "prompt": "rain hitting a metal roof, slow zoom out reveals a quiet village at dusk",
    "duration": 12,
    "generate_audio": true
  }'

# Video → video style transfer
curl -X POST 'https://api.acedata.cloud/kling/videos' \
  ... \
  -d '{
    "model": "kling-v3-omni",
    "action": "text2video",
    "prompt": "turn <<<video_1>>> into a hand-painted animation while preserving motion",
    "duration": 5,
    "mode": "std",
    "video_list": [
      {
        "video_url": "https://cdn.acedata.cloud/assets/examples/gemini/04a043bd-6b23-4b4e-945c-ce48158c3eee-3a89912507c7.mp4",
        "refer_type": "base",
        "keep_original_sound": "no"
      }
    ]
  }'

Use Cases

  • Audio-visual integrated advertising shorts —— v3 / v2-6 directly generate videos with sound effects, eliminating the need for dubbing
  • Flexible-duration content —— v3 supports any duration from 3–15 seconds, without having to change edits just to “fit 5s/10s”
  • Video remixing / style transfer —— v3-omni’s video reference input enables large-scale production of remix content
  • Branded short-video series —— kling-o1 multi-image references maintain consistency across characters and scenes