Upload a PDF, pick a parser, optionally enable chart-VLM. The same call works via POST /extract with your bearer token — see API docs.
Calls POST /parse — returns Unsiloed-shape JSON + per-segment markdown. Chart-VLM only applies to the Structured tab.
Fill a JSON schema from a PDF — the omega equivalent of Reducto's "/extract" or Unsiloed's "structured fields." Either upload a fresh PDF or pick an already-parsed job to skip the parse cost.