Topaz Video Upscaler

ID=$(curl -s -X POST https://renderhour.com/api/gateway/v1/videos/generations \
  -H "Authorization: Bearer $RENDERHOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "kie/topaz/video-upscale",
  "prompt": "Slow aerial shot over a foggy coastline at sunrise",
  "video_url": "https://example.com/input.mp4"
}' | jq -r '.id')

# 202 Accepted — the render is queued; poll until it finishes.
while :; do
  JOB=$(curl -s https://renderhour.com/api/gateway/v1/videos/generations/$ID \
    -H "Authorization: Bearer $RENDERHOUR_API_KEY")
  case "$(echo "$JOB" | jq -r '.status')" in
    succeeded) echo "$JOB" | jq -r '.data[0].url'; break ;;
    failed|cancelled) echo "$JOB" | jq -r '.error // .status'; exit 1 ;;
  esac
  sleep 5
done

fal.ai lists $0.1 per 10 seconds of output at 720p (read 2026-09-05). We do not route this endpoint yet, so that is a reference, not a quote.

Not on the gateway yet: kie/topaz/video-upscale has no price in the billing table, so this request is refused with unpriced_model rather than run. The snippet is here so the shape is clear the day it opens.

Or copy the request and run it with your RenderHour API key — the one in your profile.