YouTube bumper ad 5-6 seconds: generate at 6 or trim on Sume
YouTube's help page puts bumper ads at 5 to 6 seconds. Which Sume video models can generate a 6 second clip directly, and when a $0.02 trim is the better route.

For a YouTube bumper ad, set duration to 6 in a POST /v1/videos body on any model whose catalog range includes 6 seconds, or generate a longer clip and cut it with the $0.02 video-trim route. YouTube's help page, read on 2026-10-07, gives bumper ads a length between 5 and 6 seconds. Every Sume model that generates from a prompt or image lists 6 seconds in its range, so the real choice is not whether you can, but whether a native 6 second clip or a trimmed longer clip gives you a better ad.
What YouTube's page states
The page we fetched described ad formats, not file specifications. It gave three length notes and said that aspect ratio, file format and file size were not in that article.
| YouTube format | Length stated on the page |
|---|---|
| Bumper ads | Between 5 and 6 seconds |
| Skippable in-stream ads | No max length, under 3 minutes recommended |
| YouTube Shorts ads | Under 60 seconds recommended |
Generate at 6 seconds
Native generation is the simplest path. Seedance, Kling, Wan, MiniMax and Gemini Omni all list 6 seconds in their duration range, and the video guide documents duration as an integer number of seconds. A native 6 second clip is planned as a complete beat: a model told to make six seconds will pace motion for six seconds, and so it is usually the better choice when the bumper is one idea with a start and an end.
Mind the model limits. Under sume/auto the duration is 3 to 10 seconds with a default of 8, so if you leave duration out you get an 8 second clip, which is outside a bumper. Always send duration explicitly.
curl -sS -X POST https://api.sume.com/v1/videos \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: bumper-mug-v1" \
-d '{"model":"seedance-2","prompt":"A mug is set down; logo appears","duration":6,"aspect_ratio":"16:9"}'Trim a longer clip to 6 seconds
Trimming is better when you already have a strong 10 or 15 second ad and want its best six seconds as a bumper. The video-trim route takes video_url, start and one of end or duration, and the range is from start up to but not including end. At $0.02 per job, cutting three different six second windows from one source costs $0.06, which is far less than generating three new clips.
The tradeoff is story shape. A cut from a longer ad can end mid-motion. Pick a start that begins at a visual beat, and look at the final frame of the result before you use it.
Which route to choose
Use a native 6 second generation when the bumper is new creative. Use trim when you are repurposing a winner. In both cases keep the audio in mind: models differ on whether sound can be turned off, and Gemini Omni under sume/auto always has native audio on.
- Always send duration; do not rely on the default.
- Pick 6, not 5, if you want room for a closing beat inside the stated 5 to 6 seconds.
- Verify length on the output with a probe before you upload.
What is not covered
The YouTube page we read did not give aspect ratios or file limits, so check those on Google's current requirements before upload. Our claims about Sume are limited to the catalog durations and the trim price in the Sume docs.
Cost shape of a bumper set
Suppose you want four bumper variants with different openings. Generating them natively is four jobs, and you read the price of each from usage.cost after the first completes. Trimming four six second windows out of one existing ad is four trim jobs at $0.02 each, so $0.08 in total, and no new generation at all.
A mixed approach often works best: generate one longer master with a strong body, cut four bumper windows from it, and only regenerate the window that does not cut cleanly. Keep a sheet that records which window of which master each bumper came from, so that a winner can be rebuilt as a full ad.
Before upload, run the clip through a probe and confirm that the duration reads between 5 and 6 seconds. The video-inspect route returns probe facts for a media.sume.com clip, and the trim result itself reports duration_seconds and the actual start. If the duration is a fraction over 6 seconds, shorten the trim by a tenth of a second and rerun it with a new idempotency key version.
Sources
Related posts
More in Use cases
- YouTube Shorts ad under 60 seconds: stitch clips with Timeline
YouTube suggests Shorts ads under 60 seconds. Sume Timeline 1.0 stitches clips into one MP4 at $0.10 per output minute, so 60 seconds costs $0.10.
- AI album cover generator: square art at 3000×3000
Generate square album art, then upscale: Apple recommends at least 3000×3000. On Sume, generate 2400×2400 and upscale it 1.25× to reach 3000×3000.
- AI avatar for online course videos: build and update lessons
Use an AI avatar as your online course instructor: one reusable avatar, a short talking video per section, captions, and one Timeline join per lesson.
- Talking avatar for PowerPoint presentations, slide by slide
Make a talking avatar presenter for PowerPoint: one Sume clip per slide, up to 60 seconds each, in 16:9 or 4:3 to match the slide, inserted as MP4.
Written by Sume