v10

latestOpenAPI 3.1.0raw.githubusercontent.com2026-07-0643140398.0 KB

Artifacts

Ingest Content

Unified endpoint to ingest files, text, URIs, or already processed content

post/v1/artifacts/ingest

Request body

artifactstring required

Unique identifier for the text being ingested within the collection

collectionstring

Collection name to group related documents for better organization and filtering

metadataobject nullable

Optional metadata dictionary containing additional document information. If provided, must include 'file_name' with a valid file extension

Example request

{
  "artifact": "company_mission_statement",
  "collection": "corporate_docs",
  "input": {
    "type": "text",
    "value": "Our mission is to democratize access to artificial intelligence through secure, private, and user-controlled AI systems that respect data sovereignty and privacy."
  },
  "metadata": {
    "author": "CEO",
    "created_date": "2024-01-01",
    "department": "Executive",
    "file_name": "mission_statement.txt"
  }
}

Response

Content successfully ingested and processed

object'list'

Response object type, always 'list' for ingestion responses

model'private-gpt'

Model identifier, always 'private-gpt'

Example response

{
  "data": [
    {
      "artifact": "quarterly_report_q1",
      "doc_metadata": {
        "department": "finance",
        "file_name": "Q1_Report.pdf",
        "page_number": 1,
        "total_pages": 15
      },
      "object": "ingest.document"
    }
  ],
  "model": "private-gpt",
  "object": "list"
}