Digest one document into a bounded summary with verified quotes.

Send the document as ordered chunks. Each quote in the result is an exact span of the chunks it cites. A repeat of an identical request (same chunks, focus and digest_version) returns the stored digest at no charge.

Answers 200 with status: ready when the digest is available within wait_s, otherwise 202 with status: running; poll GET /v2/insights/digests/{digest_key}. A digest that stopped before finishing reports failed with code interrupted: send this request again, since document text is never stored. The result is kept for ttl_days. A successful digest is billed once; a failed one is not billed.

post/v2/insights/digests

Request

  • Base URL
    https://api.imentiv.ai (the document declares no server; this is the origin it was published from)
  • Auth
    • API key in header X-API-Key
    • or HTTP bearer

Request body

productstring nullable
wait_snumber

How long to wait for a new result before answering 202 running.

ttl_daysinteger

How long a finished result is kept.

digest_versionstring required

Response

Successful Response

status'ready' | 'running' | 'failed' required
digest_keystring required
digestobject nullable
cachedboolean
expires_atstring date-time nullable

Changes