POST /api/v1/deidentify/text/proprietary/outputs/pdf

De-identifies a PDF file word by word using Guardian Layer, preserving page layout. Sensitive words are replaced using PDF redaction annotations. Returns a de-identified PDF.

note

This endpoint processes PDFs with a text layer. Scanned PDFs without selectable text will not have content de-identified. For image-based documents, use the image endpoint instead.

---

Request

POST /api/v1/deidentify/text/proprietary/outputs/pdf
X-API-Key: cai_your_key_here
Content-Type: multipart/form-data

Form fields

Field	Type	Required	Default	Description
file	file	Yes	—	A .pdf file
domain	string	No	General	General, Medical, Finance, or Custom
masking_type	string	No	transform	redact fills changed words with a black rectangle; transform replaces with alternative text
pii_entities	string	No	all	Comma-separated entity types. Leave empty for all

---

Response

Returns application/pdf. The filename is prefixed with deid_.

---

Example

import requests

with open("contract.pdf", "rb") as f:
    response = requests.post(
        "https://api.custodianai.com/api/v1/deidentify/text/proprietary/outputs/pdf",
        headers={"X-API-Key": "cai_your_key_here"},
        files={"file": ("contract.pdf", f, "application/pdf")},
        data={
            "domain": "Finance",
            "masking_type": "redact",
            "pii_entities": "PERSON,ID_NUMBER",
        },
    )

with open("deid_contract.pdf", "wb") as out:
    out.write(response.content)

---

Error responses

Status	Description
400	File is not a .pdf or is malformed
401	Missing or invalid API key
403	Key expired or character limit reached
500	PDF processing library unavailable