Image-to-Video
AI-generatedSummary
Generate a video from an input image URL using the selected Kling image-to-video model and optional guidance parameters.
Inputs
- Model (required) — Select the Kling model version to use for the image-to-video generation.
- Image URL (required) — URL of the input image to transform into a video.
- Prompt — Optional text prompt to guide the image-to-video generation.
- Duration — Length of the generated video in seconds (5, 10, or 15 for Kling 3.0).
- Aspect Ratio — Desired aspect ratio of the output video (16:9, 9:16, 1:1).
- Generation Mode — Generation mode for Kling 3.0 (Standard or Pro for higher resolution), ignored for older models.
- Enable Sound Effects — Enable native audio/sound effects to add richer audio in Kling 3.0 models.
- CFG Scale — Controls creativity versus adherence to the prompt (range 0 to 1); used by older models.
- Seed — Seed number for randomization; 0 means random seed.
- Negative Prompt — Elements or features to avoid in the generated video.
- Tail Frame URL — Optional URL of an image to append as the end frame of the video.
- Reply URL — Webhook URL called upon task completion.
- Reply Ref — Custom reference string included in the webhook callback.
- Captcha Token — reCAPTCHA v3 Enterprise token; leave empty for kie.ai to handle it.
- Wait for Completion — Whether to wait and poll for the task to complete before returning the result.
Output shape
a list of JSON objects representing the API response, including task status and result URLs for the generated video if waiting for completion; otherwise the task creation response containing the task ID.
If 'Wait for Completion' is true, the node polls the task status every 3 seconds (up to 5 minutes) and returns the completed task result. The response includes parsed fields such as 'resultUrls' or 'videoUrls' for direct access to generated content URLs. If false, the node returns immediately after task creation with the task ID and metadata.