detach_source_has_no_audio: check for an audio track first

detach_source_has_no_audio means the video you asked Sume to detach audio from has no audio track, so there is nothing to extract. Avoid it by reading probe.has_audio first with POST /v1/video-inspect and frames: false, which returns the probe alone.

This page follows the Audio detach and Video inspect docs, read 2026-09-29.

Where does the error come from?

It is raised by the worker, after the job is admitted, not at submit. The neighboring refusals are below.

How do I check has_audio without paying for stills?

Send a video inspect with frames: false. The docs say probe and stills are unbilled, and that this is enough to read probe.has_audio. Inspect defaults to mode: sync, waiting up to 30 seconds.

curl -X POST https://api.sume.com/v1/video-inspect \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: inspect-probe-001" \
  -d '{"video_url":"https://media.sume.com/artifacts/artf_demo/talk.mp4","frames":false}'

What does a successful detach return?

A new audio artifact: audio_url, duration_seconds, format, channels, sample_rate (null when inherited), source_duration_seconds and optional range and warnings[]. The default format is sample-exact wav; mp3 at 128 kbps is the other option. The video is untouched.

Does a silent clip block the rest of my pipeline?

Only the audio step. Inspect's transcript has its own guard, inspect_source_has_no_audio, so branch on probe.has_audio once and skip both. To add sound to a silent video, see checking for silent audio before adding music.