A 45-second dance reference needs two Kling motion control jobs: $7.09
Sume's Kling 3.0 motion control caps one reference at 30 seconds. A 45-second dance is a 30 s plus a 15 s job, about $7.09, joined by a timeline.

A 45-second dance does not fit in one Kling 3.0 Motion Control job on Sume, because motion_video_url and duration_seconds both cap at 30 seconds. Cut the dance into a 30-second part and a 15-second part, run two jobs with the same still, and join the results on a timeline. At $0.1575 per second that is $4.725 plus $2.3625, about $7.09 in total.
Kling's own guide gives the same ceiling: action videos run 3 to 30 seconds (read 2026-10-07). The split is a property of the model, not of Sume.
Where to cut
Cut on a beat where the dancer's pose is close to neutral. The second job starts from the same still, so each part is generated on its own. A cut mid-spin is likely to show as a jump.
- Prefer a cut at a held pose or a step, not during a turn.
- Make the two parts roughly 30 and 15 seconds, or 25 and 20. Both parts must be at least 3 seconds by Kling's rule.
- Keep the same still, the same
character_orientation, and the same prompt in both jobs, so the character looks the same. - Write down the exact lengths, since you declare them as
duration_secondsfor each part.
The cost, job by job
Each job is reserved at admit and captured on completion. The two jobs are independent, so if the second fails you pay only for the first and can retry the second.
| Job | Reference seconds | Rate per second | Cost |
|---|---|---|---|
| Part 1 | 30 | $0.1575 | $4.725 |
| Part 2 | 15 | $0.1575 | $2.3625 |
| Total | 45 | $0.1575 | $7.0875 (about $7.09) |
Trim, submit, join
If the dance clip is already on Sume, POST /v1/video-trim cuts a [start, end) range into a new MP4 for $0.02 per job, and returns a media.sume.com URL you can use as the motion_video_url. Part 1 is start 0 with a duration of 30, and part 2 is start 30 with a duration of 15.
Submit both motion-control jobs with different idempotency keys, then place the two finished clips back to back on a timeline. The audio of each part is kept by default, so the music runs through the joint if the trim points are exact. Use precision: exact on the trim so the parts line up frame for frame.
curl -X POST https://api.sume.com/v1/kling/3.0/motion-control \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: dance-part-1" \
-d '{"image_url": "https://example.com/mascot.png", "motion_video_url": "https://example.com/dance-part-1.mp4", "duration_seconds": 30, "character_orientation": "video", "mode": "async"}'What to expect at the joint
Two separate generations will not match perfectly at the cut: the face, clothing, and lighting can drift a little. The docs promise no continuity between jobs. Check the join frame by frame before you publish. For the 60-second version of this plan, read two 30-second jobs and a timeline join.
Sources
Related posts
More in Media tools
- Audio detach size for a 180 s Short: 34.56 MB wav, 2.88 MB mp3
The byte arithmetic for a 3-minute track from audio detach: stereo 48 kHz wav, mono 16 kHz wav and 128 kbps mp3, with the fields that set each size.
- Can Gemini Omni edit change a video's aspect ratio? No
On Sume, Omni's edit mode takes a video_url and a prompt, defaults to 720p and accepts no aspect_ratio or duration. A 16:9 source stays 16:9, so plan the crop.
- Caption a talking-head video: style, placement and phrasing
Caption a talking-head clip with the Sume API: pick a style, move the line off the face, set words per card. One request, $0.20 for up to 60 seconds.
- Captions for a video with loud background music: script text or cues
Music can bury speech and trip speech-to-text. Sume documents three ways to supply your own wording: script_text, a words array or cues. Which to pick.
Written by Sume