Sora 2 Video Generation API Guide

Sora 2 Overview

Sora 2 is OpenAI's video generation model available through Ruxa AI. It supports text-to-video, image-to-video, and character animation features.

Available Models

Model	Resolution	Duration
`sora-2`	1280x720, 720x1280	10s, 15s
`sora-2-hd`	1792x1024, 1024x1792, 1280x720, 720x1280	10s, 15s, 25s

Basic Text-to-Video

Generate a video from a text prompt:

curl --request POST \
  --url https://api.ruxa.ai/api/v1/tasks/create \
  --header 'Authorization: Bearer sk-xxxxxx' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "sora-2",
    "input": {
      "prompt": "A golden retriever running through a sunlit meadow, slow motion, cinematic quality",
      "seconds": "10",
      "size": "1280x720"
    }
  }'

Image-to-Video

Animate a reference image:

curl --request POST \
  --url https://api.ruxa.ai/api/v1/tasks/create \
  --header 'Authorization: Bearer sk-xxxxxx' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "sora-2",
    "input": {
      "prompt": "The person in the image starts walking forward",
      "input_reference": "https://example.com/your-image.jpg",
      "seconds": "10",
      "size": "1280x720"
    }
  }'

Character Animation

Create reusable characters for consistent video generation:

curl --request POST \
  --url https://api.ruxa.ai/api/v1/tasks/create \
  --header 'Authorization: Bearer sk-xxxxxx' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "sora-2",
    "input": {
      "prompt": "The character waves hello",
      "character_url": "https://example.com/character-video.mp4",
      "character_timestamps": "1,3",
      "character_create": true,
      "seconds": "10"
    }
  }'

Note: Character videos cannot contain real people.

Style Options

Apply visual styles to your videos:

Style	Description
`winter`	Winter/snow aesthetic
`comic`	Comic book style
`news`	News broadcast style

{
  "model": "sora-2",
  "input": {
    "prompt": "A cat playing with yarn",
    "style": "comic",
    "seconds": "10"
  }
}

Parameters Reference

Parameter	Type	Required	Description
`prompt`	string	Yes	Video description
`seconds`	string	No	Duration: "10", "15", "25" (HD only)
`size`	string	No	Resolution (see table above)
`input_reference`	file/url	No	Reference image for image-to-video
`watermark`	boolean	No	Add watermark (default: false)
`private`	boolean	No	Hide video from public (default: false)
`style`	string	No	Visual style
`character_url`	string	No	Video URL for character creation
`character_timestamps`	string	No	Character appearance range "start,end" (1-3s)
`character_create`	boolean	No	Auto-create character from result
`character_from_task`	string	No	Use character from existing task

Query Task Status

curl --request GET \
  --url 'https://api.ruxa.ai/api/v1/tasks/query/{taskId}' \
  --header 'Authorization: Bearer sk-xxxxxx'

Prompt Tips

For best results:

Be specific - Describe camera movements, lighting, and mood
Use cinematic terms - "dolly shot", "tracking shot", "close-up"
Match complexity to duration - Longer videos can handle more complex scenes
Include style descriptors - "cinematic", "slow motion", "professional"

Example prompt:

"A cinematic shot of ocean waves crashing against rocky cliffs at sunset, golden hour lighting, slow motion, aerial drone perspective, professional film quality"