Image to Video
AI-generatedSummary
Generate a video from a given image URL or reference images using the Veo 3.1 model by Kie.ai, with options to customize the model variant, aspect ratio, generation mode, duration, watermark, and other parameters.
Inputs
- Model — Select the Veo 3.1 model variant to use: Standard, Fast, or Lite.
- Prompt (required) — Enter text prompt to guide the video generation from the image.
- Image URL (required) — Specify the URL of the source image to generate the video from (ignored if Reference Images are set).
- Reference Images — Optional multiple reference image URLs that override the Image URL to guide the video generation.
- End Frame URL — Optional URL of an end frame image to be used for image-to-video generation.
- Aspect Ratio — Choose the aspect ratio of the resulting video: 16:9, 9:16, or Auto.
- Generation Type — Specify the explicit generation mode: Auto (let Kie infer), Text to Video, First and Last Frames to Video, or Reference to Video.
- Watermark — Optional text to add as a watermark on the generated video.
- Duration (Seconds) — Set the length of the generated video, either 5 or 8 seconds.
- Enable Translation — Enable translation feature if needed (boolean).
- Seed — Numeric seed for deterministic generation; set 0 for random seed.
- Reply URL — Optional webhook URL to notify when the video generation task completes.
- Reply Ref — Optional custom reference string passed back in the webhook callback.
- Wait for Completion — If true, the node polls until the video generation completes (up to 5 minutes). If false, returns immediately with task info.
Output shape
a single JSON object containing task status and result URLs
The output includes the final video generation task data, such as video URLs and status information. If waiting for completion is enabled, the object reflects the completed task; otherwise, it returns the initial task creation response. Error responses may be returned if the generation fails.