16:9 clip in Snapchat's 720x1280 frame: contain fills 31.6 percent

If you drop a 1280x720 clip into Snapchat's 9:16, 720x1280 frame with fit contain, the picture is 720 pixels wide and 405 pixels tall, so it fills 31.6 percent of the height and leaves 875 empty rows. Fit cover fills the frame but keeps only 31.6 percent of the width. Fit blur keeps the whole picture and fills the empty rows with a blurred copy.

Snapchat's frame

Snapchat's ad format page lists 9:16 and 720x1280 for single image or video ads, story ads, collection ads and Commercials, and 3 to 180 seconds for the first three.

The geometry of each fit mode

Timeline 1.0 slots take fit set to cover (default), contain, stretch or blur (Sume docs, read 2026-10-08). The numbers for a 1280x720 source and a 720x1280 output:

Fit modes for 1280x720 into 720x1280, computed 2026-10-09, as listed below.

• contain: scale 720 / 1280 = 0.5625; result 720 x 405 picture, 875 rows empty (68.4%)
• cover: scale 1280 / 720 = 1.778, picture 2,276 x 1,280; result 720 of 2,276 columns kept (31.6%), sides cropped
• stretch: scale Both axes fitted; result Whole picture, distorted
• blur: scale Contain plus blurred fill; result Whole picture, no empty rows

Which one to ship

Pick blur when the whole frame matters, such as a wide product shot or text near the edges. Pick cover only when the subject sits in the middle 31.6 percent of the width. Stretch changes shapes, so keep it for abstract footage.

A 16:9 clip also leaves the top 437 rows and the bottom 438 rows for blur ((1280 - 405) / 2 = 437.5). Any burned-in captions or logos should go in the picture area or be placed deliberately in the blurred bands.

• Timeline render: $0.10 per output minute, rounded up.
• A 30-second ad is 1 minute, so 1 x $0.10 = $0.10.
• The plan endpoint is an unbilled preflight.

Checking the result

After the render, extract a still with video frames and look at the top and bottom of the frame. In blur mode the fill should look like a soft copy of the picture, not a flat color. If the fill is distracting, a cover render of a tighter crop may serve better.

Also run video inspect with frames false and confirm width 720 and height 1280 in the probe. A frame that is 1280x720 means the output size was not set and the render used the source shape.

Text placement

Snapchat's page does not give a safe zone for these formats, so do not assume one. In a contain or blur render, the picture occupies the middle 405 rows (rows 438 to 843), and any text that must stay readable on the picture belongs there. Text placed in the blurred bands is legible but sits on a moving background, so keep it short.

Neither Snapchat nor Sume enforces taste here. The page lists the frame and the length, and the timeline accepts any of the four fit modes, so the choice of how a wide picture sits in a tall frame is yours. Render the two likely options on a 10-second test, compare the stills side by side, and use the one that keeps the subject readable. The unbilled plan endpoint confirms the document compiles before you spend anything.