Pitch sizzle reel: trim three episodes and join them for $0.19

A 60 second pitch reel from three finished episodes costs $0.19 on Sume: three video-trim jobs at $0.02 each, three audio-detach jobs at $0.01 each and one Timeline 1.0 render at $0.10 for one started minute. The reel is the thing you attach when a program such as The Next Episode, reported by Metricool as funding creator-led series, asks to see your work.

The method below takes 20 second best moments from each episode, keeps their original sound in order and adds short fades between them.

The cost, line by line

Every step is worker ffmpeg, with no model inference, so the prices are flat per job.

Step 1: trim each episode

Video trim takes a start and exactly one of end or duration, between 0.2 and 900 seconds. The default precision is exact, a frame-accurate re-encode, which is what you want when the cut lands on a line of dialogue.

curl -X POST https://api.sume.com/v1/video-trim \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: reel-trim-ep1" \
  -d '{"video_url":"https://media.sume.com/artifacts/artf_ep1/ep1.mp4","start":12,"duration":20}'

Step 2: detach the sound

The timeline takes its sound from audio.url or from audio.parts[], up to 20 parts. Run audio-detach on each trimmed clip to get a wav. A clip with no audio track fails with detach_source_has_no_audio, so check silent clips first with the inspect endpoint.

The three detached wavs become the parts of the timeline audio, in the same order as the video slots.

Step 3: render the reel

Each video slot needs source_url, start and duration. The first slot must start at 0. A transition is not allowed on the first slot and cannot exceed one second.

{
  "audio": {"duration_seconds": 60, "parts": [
    {"url": "https://media.sume.com/artifacts/artf_a1/ep1.wav"},
    {"url": "https://media.sume.com/artifacts/artf_a2/ep2.wav"},
    {"url": "https://media.sume.com/artifacts/artf_a3/ep3.wav"}]},
  "video": [
    {"source_url": "https://media.sume.com/artifacts/artf_t1/ep1.mp4", "start": 0, "duration": 20},
    {"source_url": "https://media.sume.com/artifacts/artf_t2/ep2.mp4", "start": 20, "duration": 20,
     "transition": {"type": "fade", "duration": 0.25}},
    {"source_url": "https://media.sume.com/artifacts/artf_t3/ep3.mp4", "start": 40, "duration": 20,
     "transition": {"type": "fade", "duration": 0.25}}]
}

Check before you pay and what to avoid

Post the same body to /v1/timeline-1.0/plan first. It is unbilled and returns the duration, segment count and estimated cost. If the parts are shorter than the declared duration the render is refused with audio_parts_shorter_than_duration, so make each wav at least 20 seconds. The default output is 1080x1920, which suits vertical episodes.

Sume does not decide which moments are best. Pick the in-points yourself from your own cuts.