X standalone video lists 16:9 and 1:1: set the Sume render size

Sume's timeline render outputs 1080 by 1920 (9:16) unless you say otherwise, and X's standalone video ad spec lists 16:9 and 1:1. So set output.width and output.height in the render body: 1920 by 1080 for landscape or 1200 by 1200 for square, both of which X names as recommended sizes.

What X lists

X's creative specs page gives the recommended video size for standalone video ads as 1200 by 1200 for 1:1 and 1920 by 1080 for 16:9, and says larger videos are optimized for when users click to expand. It notes the legacy composer supports any ratio between 2:1 and 1:1. Ads with website buttons recommend 800 by 450 (16:9) or 800 by 800 (1:1).

The Sume request

The timeline output block accepts width and height as even integers from 256 to 2160, so both sizes are valid. A body for a 20-second landscape render looks like this, using files already imported to your workspace.

The default fit for each slot is cover, which crops to fill the frame. A 9:16 source placed in a 16:9 frame loses most of its height under cover. Use fit: blur to keep the whole clip in the middle with a blurred copy behind, or contain to leave bars. The fit options are cover, contain, stretch and blur.

{
  "audio": {"url": "https://media.sume.com/artifacts/artf_demo/voice.wav", "duration_seconds": 20},
  "video": [
    {"source_url": "https://media.sume.com/artifacts/artf_demo/clip.mp4",
     "start": 0, "duration": 20, "fit": "blur"}
  ],
  "output": {"width": 1920, "height": 1080, "fps": 30}
}

Price and limits

A 20-second render bills one minute at $0.10 (Sume catalog, read 2026-10-08). The plan endpoint, POST /v1/timeline-1.0/plan, is unbilled and returns duration_seconds, billable_minutes and an estimated cost before you commit. It does not download media.

Pick 1:1 when you want one file for several places. X lists 1:1, and TikTok's non-Spark ads accept 1:1 at 640 by 640 or more, according to TikTok's own spec page read the same day. Snapchat's ad formats list only 9:16, so square is not a universal answer.

Check what came back

A render result has video_url, duration_seconds, segment_count and billable_minutes. To confirm the frame size, send the new video_url to video inspect with frames set to false: the probe returns width, height, fps, video_codec and size_bytes. If width and height are not what you set, do not upload; resubmit with the output block present.

Frame rate matters too. If you omit output.fps the job follows the sources, and if you set a rate different from the source, Sume repeats or drops frames and reports output_fps_resamples_sources. X accepts 29.97 or 30 FPS and up to 60, so set 30 explicitly for a mixed set of clips instead of leaving it to the source.

• Square 1200x1200 keeps one file for X, TikTok non-Spark and LinkedIn, whose video page lists 1:1 at 360x360 to 1920x1920.
• Landscape 1920x1080 sits at LinkedIn's landscape maximum and X's 16:9 recommendation.
• Vertical 1080x1920 is the Sume default, and the standalone X spec above does not list it.