API Reference

OmniVideo API Documentation

Use the Gemini Omni video model through a compact REST API for asynchronous text-to-video and image-to-video generation.

Base URL

https://omnivideoapi.com/api

Authentication

Send your API key as a Bearer token from your server-side code.

Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

Create Video Task

POST /generate

{
  "model": "gemini-omni/video",
  "input": {
    "prompt": "A cinematic product reveal shot with smooth camera motion",
    "mode": "std",
    "aspect_ratio": "16:9",
    "duration": "5",
    "sound": true
  }
}

Parameters

Field	Type	Required	Notes
model	string	Yes	Use `gemini-omni/video`.
input.prompt	string	Yes	Prompt up to 2500 characters for a single-shot task.
input.mode	string	No	`std` or `pro`. Default is `std`.
input.aspect_ratio	string	No	`16:9`, `9:16`, or `1:1`.
input.duration	string	No	Supported range: 3 to 15 seconds.
input.sound	boolean	No	Whether to request generated audio. Default is true.
input.image_urls	string[]	No	Public image URLs for image-to-video tasks.

Response

{
  "code": 200,
  "message": "success",
  "data": {
    "task_id": "task_abc123",
    "status": "IN_PROGRESS"
  }
}

Check Status

GET /status?task_id=task_abc123

Poll every 5 to 10 seconds until status becomes SUCCESS or FAILED. Successful responses include result URLs in data.response.resultUrls.