Video Generation

Video generation is asynchronous. Create a task with POST /v1/videos, retain the returned task ID, then retrieve its state until the result is ready.

The public catalogue lists the current video model, supported durations, available frame sizes, audio behavior, and starting price.

Create a task

curl "https://api.offrail.ai/v1/videos" \
  -H "Authorization: Bearer $OFFRAIL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2-5-uncensored",
    "prompt": "Slow dolly through a neon-lit studio, shallow focus",
    "seconds": 8
  }'

{
	"id": "video_123",
	"status": "queued"
}

Retrieve the task

curl "https://api.offrail.ai/v1/videos/video_123" \
  -H "Authorization: Bearer $OFFRAIL_API_KEY"

Continue polling while the status is queued or in_progress. When the task is complete, retrieve the generated content from the URL or content endpoint in the response.

API reference

• Create video
• Retrieve video
• Retrieve video content
• Routing behavior