Extract · Custom fields (IDP)
Extract Custom fields (IDP) via API — structured data from 3¢/page
PaperOffice AI extracts Custom fields (IDP) data via API using LLM-based semantic extraction combined with Surya OCR geometry. Pricing from 3¢/page (Basic); Premium and Ultra for higher model tiers. Extract custom IDP field definitions configured for the account from uploaded documents.
Betroet af førende virksomheder verden over
SUPER TRUST
The world’s exclusive sole DMS partner for QNAP and ASUSTOR.
Dedicated landing pages, direct integration, data stays on your NAS.
QNAP
Sole official DMS partner for QNAP NAS — documents stay on your device, AI runs through PaperOffice.
More
ASUSTOR
Sole official DMS partner for ASUSTOR NAS — BYOS, on-premises and full control over where data lives.
More
Extracted fields
Source: Extract custom fields (IDP) · POST https://api.paperoffice.ai/latest/job/add/workflow
| Field / description |
|---|
| Extract custom IDP field definitions configured for the account from uploaded documents. |
Code example
curl
curl -X POST "https://api.paperoffice.ai/latest/job/add/workflow" \
-H "Authorization: Bearer po_sk_xxx" \
-F "[email protected]" \
-F "model=premium"Python
import requests
api_token = "po_sk_xxx"
file_path = "document.pdf"
response = requests.post(
"https://api.paperoffice.ai/latest/job/add/workflow",
headers={"Authorization": f"Bearer {api_token}"},
files={"file": open(file_path, "rb")},
data={"model": "premium"},
)
job_result = response.json()
print(job_result)Example response
{
"status": "success",
"job_id": "job_…",
"result": {
"idp_collection": "…",
"fields": {}
}
}Truncated schema — full fields in llms-full.txt / Postman.
Pricing
| Tier | Price / page |
|---|---|
| Basic | 3 ct / €0.03 |
| Premium | 4 ct / €0.04 |
| Ultra | 10 ct / €0.10 |
API credit note: 10 credits (base, at `priority=500`)
FAQ
What does extracting Custom fields (IDP) cost?
AI DMS / IDP processing costs Basic 3¢, Premium 4¢ and Ultra 10¢ per page.
Which endpoint is used?
Per API docs: POST https://api.paperoffice.ai/latest/job/add/workflow (pipeline/slug: workflow).
Is extraction available via MCP?
Yes. The same capabilities are reachable via the PaperOffice MCP server — see /mcp/.
Where is data processed?
On PaperOffice EU infrastructure on owned hardware.
