Video 1.0 metadata field: stored on the job, not sent upstream

The optional metadata field on Video 1.0 is caller metadata stored on the job; it is not sent to the provider. Use it to tag your own job records, not to steer the model.

What does the docs table say?

metadata is described as caller metadata stored on the job, not sent to the provider. The docs show no size limit or schema for it here, so keep the object small and plain.

Which fields go upstream and which stay with Sume?

The Video 1.0 fields split by what they control.

How do I use it?

Add your own reference, such as an order or batch label, in the request body.

curl -X POST https://api.sume.com/v1/video-1.0/generate \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: v1-meta-001" \
  -d '{"prompt":"A clean product spin on a turntable","duration":5,"metadata":{"batch":"batch_demo"}}'

Where do I read it back?

Job status and result are read with GET /v1/jobs/{id}/status and /result; see the Jobs and results docs for the fields those return. The docs here do not say the metadata is echoed in a specific field, so check a real response.

How do I keep my own records in sync?

Store the job id from the submit response so you can recover work after a restart. The jobs page says to poll with exponential backoff and stop on completed, failed or canceled.

Pair your own tag in metadata with the returned job id and the Idempotency-Key you sent, so a retry maps back to the same record.