Actions6
- Chat Actions
- Audio Actions
- Video Actions
- Image Actions
Video → Generate Video
AI-generatedSummary
Start, remix, check status, or download AI-generated videos using OpenAI Sora or Veo models.
Inputs
- videoMode (required) — Select how the node should behave: Start a job, Remix an existing video, Check status, or Download completed video.
- videoModel (required) — The video generation model to use: OpenAI Sora 2, OpenAI Sora 2 Pro, or Veo 3.0.
- prompt — Text prompt describing the video to generate. Required for Start and Remix modes.
- soraSize — Resolution of the video (e.g., 1280x720). Applies to Start mode.
- seconds — Duration of the video in seconds (4 or 8). Applies to Start mode.
- inputDataMode — For Start mode with Sora Pro, provide an input image via Binary File or Image URL.
- imageBinaryProperty — Binary property containing the input image (used when inputDataMode is Binary File).
- imageUrl — URL of the input image (used when inputDataMode is Image URL).
- videoId — OpenAI video ID returned from Start (used for Sora models in Status, Download, Remix).
- operationName — Vertex AI operation name returned from Start (used for Veo models in Status and Download).
- videoAdditionalFields — Additional fields: Sample Count, Location, Resume URL.
Output shape
Depending on the mode and model, returns JSON metadata and optionally binary video data.
Start: JSON with success flag, model, prompt, size, seconds, and videoId (Sora) or operationName (Veo). Remix: JSON with success flag, previous video ID, new prompt, and new videoId. Status: JSON with job status and, for Veo, a download URL when complete. Download: Binary MP4 video plus JSON with success flag and file size.
Examples
Example 1: Generate a Sora video from a text prompt and later download it.
Set videoMode to Start, videoModel to openai/sora-2, provide prompt, soraSize, and seconds.