Extract audio from an Instagram video: Sume audio detach for intros

Inside Edits you tap the Audio icon, then the folder icon, and extract audio from a video. Outside the app, with a video you host on Sume, POST /v1/audio-detach returns a new audio artifact (wav by default, or mp3) and leaves the video untouched.

Edits steps are from Instagram's creators blog; Sume behavior is from Audio detach, read 2026-10-01.

What does Edits offer?

Instagram says you can extract audio directly from a video or import an audio file from your device, then save it in your library to reuse across future projects, for example recurring intros or signature sounds. That saved file lives in the app.

How do I do the same thing with the API?

Send video_url for a workspace media.sume.com video (import first with POST /v1/media-imports) and an Idempotency-Key. Poll GET /v1/jobs/:id/status, then GET /v1/jobs/:id/result, which returns audio_url as a new artf_ artifact. Hosted MCP has an audio_detach tool with the flow audio_detach, jobs_wait, jobs_result.

For a short signature sound, pass range as { start, end? } in seconds. The docs describe the public rate on the page and say to confirm it live in GET /v1/catalog.

curl -X POST https://api.sume.com/v1/audio-detach \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: intro-detach-001" \
  -d '{
    "video_url": "https://media.sume.com/artifacts/artf_demo/talk.mp4",
    "range": { "start": 0, "end": 4 },
    "format": "mp3"
  }'

Which options fit a reusable intro?

What can go wrong?

A source with no audio track fails with detach_source_has_no_audio. A range whose end is not after start fails with audio_detach_range_empty. Source video is capped at 1800 s and output at 900 s. A saved intro is just an artf_ URL, so you reuse it by passing it to later requests; the overlap with trimming is covered in trim, filter and detach audio.