Video generation

• client.generateVideo()
• client.listVideos()
• client.getVideo()
• client.getVideoContent()
• client.getVideoDownloadUrl()
• client.cancelVideo()
• client.deleteVideo()
• client.listVideoModels()
• client.getAsyncJobWebSocketUrl("video", videoId, options)
• client.videos.websocketUrl(videoId, options)
• client.videos.create/list/get/content/downloadUrl/cancel/delete/listModels()

​

Example

const job = await client.generateVideo({
  model: "openai/sora-2",
  prompt: "A calm ocean at dusk",
});

const jobs = await client.listVideos({ status: "queued,completed", limit: 2 });

const status = await client.getVideo(job.id);
const websocketUrl = client.videos.websocketUrl(job.id, { intervalMs: 1500 });

if (status.status === "completed" && status.download_url) {
  console.log(status.download_url);
}

​

Key parameters

• model (required): Video-capable model id.
• prompt (required): Text description of the video.
• webhook (optional): Receive lifecycle notifications instead of polling only.
• output.access (optional): Control whether the job returns authenticated byte URLs, signed download URLs, or both.
• Optional provider-specific controls (for example duration, aspect, audio, or quality) when the upstream supports them.

​

Returns

{
  "id": "vid_123",
  "object": "video",
  "status": "queued",
  "polling_url": "/v1/videos/vid_123",
  "poll_after_seconds": 20
}

{
  "object": "list",
  "data": [
    {
      "id": "vid_123",
      "object": "video",
      "status": "queued"
    }
  ]
}