Replace a clip's soundtrack: detach, Lyria music, re-render on Sume

Three Sume calls do it. Audio detach pulls the track out for $0.01, the Music Router makes a replacement for $0.125, and a Timeline 1.0 render lays it under the video at $0.10 per output minute, rounded up. For a 30-second clip that is $0.235.

Why detach first

• It gives you a sample-exact wav (pcm_s16le) to inspect or transcribe. The video is untouched.
• If a reference ingest reports audio.silent true, the docs say to plan new music and discard the source audio.
• The original sound may be speech you want to keep. Detach, then transcribe, before you decide to replace it.

The cost

Rights, plainly

Swapping the sound removes the audio, not the claim on what you filmed. If the clip belongs to someone else, a new track does not make the picture yours. Use footage your workspace owns; Sume's tools read only media you imported to media.sume.com.

Detach and generate

curl -X POST https://api.sume.com/v1/audio-detach \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: detach-1" \
  -d '{"video_url":"https://media.sume.com/artifacts/artf_demo/talk.mp4"}'

curl -X POST https://api.sume.com/v1/music-router/generate \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: music-1" \
  -d '{"prompt":"Upbeat acoustic bed, a 30-second track. Instrumental, no vocals."}'

Read audio_url from the detach result and the audio artifact from the music result. Pass the music URL as audio.url on the timeline render.