deAPI icon

deAPI

Generate images, videos, transcribe audio and boost prompts using deAPI AI models

Actions12

Video → Generate a Video

AI-generated

Summary

Generates video from a text prompt or image(s) using LTX video models.

Inputs

  • source (required) — Select 'Text Prompt' for pure text-to-video or 'Image(s)' to generate video from image frames (first frame required, last frame optional).
  • prompt (required) — A text description of the desired video.
  • firstFrame (required) — The name of the binary field containing the first frame image (required when source is 'Image(s)').
  • model (required) — Select the AI model: 'LTX-Video-0.9.8 13B', 'LTX-2 19B Distilled FP8', or 'LTX-2.3 22B Distilled INT8'.
  • ratio (required) — Aspect ratio: square, landscape, or portrait.
  • options — Optional parameters including 'lastFrame', 'Ltx1Frames'/'Ltx2Frames', 'negativePrompt', resolution presets, 'seed', and 'waitTimeout'.

Output shape

The input item(s) are passed through unchanged; the video generation occurs asynchronously via a webhook, and the resulting video file is returned in the workflow's binary data (e.g., 'data') upon completion.

The operation uses n8n's 'wait' mechanism (putExecutionToWait) with a configured timeout. The actual video generation happens on the external API; n8n polls or resumes based on the webhook response.

Examples

Example 1: Text-to-video

Set Source to 'Text Prompt', provide a descriptive prompt, select a model (e.g., LTX-2.3 22B), and choose an aspect ratio (e.g., Landscape).

Example 2: Image-to-video

Set Source to 'Image(s)', provide a binary field for the first frame, optionally provide a last frame, and configure prompt, model, and ratio.

Example 3: Custom resolution

Select a model and ratio, then use the 'Resolution' options in the 'Options' collection to specify a custom width and height (e.g., 1024x576).

Links

Discussion