Story Layout holds 6 photos; Sume compose holds a still and a clip

Instagram's Layout tool can "split a single story frame into a grid of up to 6 photos", according to Meta's Stories guide. Sume's timeline compose does something narrower: it puts one still and one video on screen at the same time and returns one MP4. If your need is a 6-up photo grid, that is an app task, or a still you prepare yourself. If your need is a banner plus a clip, or a plate over a clip, compose is the right tool.

What is Layout, exactly?

The Stories guide separates Layout from Collage. Layout is the grid: one frame split into up to 6 photos. Collage is the AI tool that animates 5 or more photos. Meta's guide does not give pixel sizes for the grid cells, so I will not either.

What does compose do?

Compose takes one image.url that must probe as a still and one video.url that must probe as video. Both must already be on media.sume.com. operation is stack (two regions of one frame) or overlay (the still as a plate over the video). The output length always comes from the video; the still is held for the whole clip. Ceiling is 300 seconds.

How do I build a banner above a clip?

The defaults are a horizontal split with the still on top and a ratio of 0.5, which the docs call a half-banner. This request puts a banner image over the top 30 percent of a 720x1280 frame and the clip underneath. The layout.ratio is the still's share, and the video takes the exact remainder.

curl -X POST https://api.sume.com/v1/timeline-1.0/compose \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: compose-banner-001" \
  -d '{
    "operation": "stack",
    "image": { "url": "https://media.sume.com/artifacts/artf_demo/banner.png" },
    "video": { "url": "https://media.sume.com/artifacts/artf_demo/talk.mp4" },
    "layout": { "split": "horizontal", "image_region": "top", "ratio": 0.3 },
    "output": { "width": 720, "height": 1280 }
  }'

What if I still want a photo grid?

Make the grid first and treat it as one still. Sume's image endpoints return up to 10 images per request with n, and ratios such as 9:16 and 4:5, but nothing in the docs says they compose a grid for you, so arrange the cells in an image tool of your choice. Then pass the finished grid as image.url and put a short clip below it with compose. A mute clip only warns (compose_video_has_no_audio); the soundtrack comes from Timeline when you assemble.

Which one should I use?

Use Layout in the app for a quick 6-photo Story with no files to manage. Use compose when you need the same banner-plus-clip look on many files, or a repeatable template, because every setting is a field in a request. Check the result frame by frame before posting, since a stack keeps text and faces only where the layout puts them.