Grok Imagine Video 1.5 on Sume: image-in only, silent, $0.19 for 15 s

grok-imagine-video-1.5 on Sume is an image-to-video row: it requires a start image (image_url or first_frame_url), produces silent clips of 4 to 15 seconds at 480p or 720p, and takes no end frame and no reference media. Its Sume price is $0.0125 per second, so a 15-second clip is 15 x $0.0125 = $0.1875, shown as $0.19.

What the row accepts

The constraints come from the Video Router catalog. The list is short on purpose; the row is for cheap animation of a still, not for scene control.

How it compares on price

At $0.0125 per second the row is a fifth of Wan 3.0 at 480p ($0.0625). For a 10-second clip that is $0.125 against $0.625. The price is the same at 480p and 720p in the Sume price table, so 720p is the sensible default.

You trade away control. There is no last frame to land on, no reference image to hold a character, and no sound. Add sound afterwards with the Timeline API, or pick a row with native audio such as MiniMax H3.

• Use for: loops, product stills coming alive, social bumpers.
• Avoid for: scenes that must end on a specific frame, or that need a character held across shots.
• Because aspect ratio follows the image, crop the still to 9:16 or 1:1 before you send it.

Request

Use the Video Router fields for this row, with the image on image_url.

curl -X POST https://api.sume.com/v1/video-router/generate \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: grok-loop-001" \
  -d '{
    "model": "grok-imagine-video-1.5",
    "prompt": "Steam curls up from the cup, the light slowly shifts",
    "image_url": "https://example.com/cup.png",
    "resolution": "720p",
    "duration": 6,
    "mode": "async"
  }'

Why use a row this limited

The point is cost per looping asset. A catalogue of 200 product stills animated for 6 seconds each is 200 x 6 x $0.0125 = $15.00. The same catalogue on Wan 3.0 at 480p is 200 x 6 x $0.0625 = $75.00. If the output is a hover animation or a feed bumper, the cheaper row does the job.

Check the first output before launching the batch. With no aspect ratio field, a badly cropped still produces a badly cropped clip.