A Text to Video API With 46 Video Models on One Key

Models that fit

Every video model in the live catalog accepts a text-only prompt through generate-media — no image or reference input required.

Try it

This request uses the current default settings for a matching model and the public GenerateMediaRequest schema.

curl -X POST "https://reviral.ai/api/v1/generate-media" \
  -H "Authorization: Bearer $REVIRAL_API_KEY" \
  -H "Idempotency-Key: first-generation" \
  -H "Content-Type: application/json" \
  --data '{"model":"seedance-2","prompt":"A drone shot rising over a misty pine forest at dawn","durationSec":5,"resolution":"720p","aspectRatio":"16:9","maxCredits":106}'

Frequently asked questions

Do I need anything besides a text prompt?

No — model and prompt are the only required fields on generate-media. durationSec, resolution, and aspectRatio are optional and fall back to each model's defaults.

How long can the prompt be?

Up to 100,000 characters, though a few sentences of concrete detail — subject, action, camera move, lighting — produces more consistent results than a very long prompt.

How do I know the exact price before I render?

Read GET /api/v1/models for the current credit price per model and resolution, then pass that number as maxCredits so a live price change fails the request instead of silently overcharging you.