Cut a 15-second vertical teaser from a trailer with video-trim
After a showcase like Xbox's Tokyo Game Show 2026, studios cut teasers. One Sume video-trim call takes a 15-second range and conforms it to 1080x1920.

A showcase produces a pile of trailers and release dates in one evening. Xbox's 17 September 2026 recap lists, for example, a Call of Duty: Modern Warfare 4 release on 23 October 2026 and a Persona 4 Revival release on 18 February 2027, read 2026-10-01. A studio that already holds its own trailer in its Sume workspace can cut a short vertical teaser from it with a single video trim call. Sume does not scrape or fetch trailers from the open web, so the source must already be a media.sume.com file.
What does one video-trim call do?
It takes one clip plus start and exactly one of end or duration, and returns a new MP4 of just that range. The source is untouched. A duration can be 0.2 to 900 seconds, so a 15-second teaser is well inside the limit. The default precision is exact, a frame-accurate re-encode. keyframe is a stream copy whose cut may begin a GOP early.
How do I get 1080x1920 out of a 16:9 trailer?
Add the optional output object. It conforms width, height and fps on the way out, works only with exact, and takes width and height from 256 to 2160 and fps of 24, 25, 30 or 60. The docs do not say how a landscape picture is framed inside a vertical output, so look at the delivered clip before you publish. If you want to choose the framing, render through Timeline 1.0 instead, whose per-slot fit values are cover, contain, stretch and blur.
curl -X POST https://api.sume.com/v1/video-trim \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: trailer-teaser-001" \
-d '{
"video_url": "https://media.sume.com/artifacts/artf_demo/trailer.mp4",
"start": 42,
"duration": 15,
"precision": "exact",
"audio": "keep",
"output": { "width": 1080, "height": 1920, "fps": 30 }
}'Which parameters make the teaser?
Values from the request above and the video trim docs.
| Parameter | Value in this example | Documented range |
|---|---|---|
| start | 42 | Seconds into the source |
| duration | 15 | 0.2 to 900 seconds |
| precision | exact | exact (default) or keyword-aligned keyframe |
| output width and height | 1080 x 1920 | 256 to 2160 |
| output fps | 30 | 24, 25, 30 or 60 |
How do I find the right start time first?
Run video inspect with a frames.at list of timestamps. It returns stills as media.sume.com image artifacts, and probe and stills are unbilled; only the optional transcript is billed. Pick the second where the shot you want begins and put it in start. A request at value outside the clip's duration returns frame_time_out_of_range.
What does it cost?
The docs list $0.02 per trim job, with the source limited to 1,800 seconds. Poll GET /v1/jobs/:id/result for video_url, a new artifact. For several teasers from one trailer, send one trim per range, each with its own idempotency key.
Sources
Related posts
More in Use cases
- Gemini Media Watermark setting: visible off, SynthID stays
Gemini Apps' Media Watermark setting turns off only the visible mark. SynthID and Content Credentials stay on. India, South Korea and Vietnam need AI Ultra.
- Seamless looping food video with AI: same first and last frame
To loop a food clip, send one image as image_url and end_image_url to Gemini Omni Flash 1.1, which Google says suits seamless loops. Clips run 3 to 10 seconds.
- Gemini TTS two-speaker limit: three-voice dialogue with Sume concat
Gemini TTS configures two speakers per request. For three voices on Sume, make one TTS job per voice and join up to 20 parts with Timeline audio concat.
- GitHub social preview 1280x640: generate it with an image API
GitHub's social preview is 1280x640 PNG, JPG or GIF under 1 MB. 1280x640 passes GPT Image 2.5's size rules on Sume; here is the request and the file-size step.
Written by Sume