> ## Documentation Index
> Fetch the complete documentation index at: https://docs.formepdf.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Redact PDF

> Redact sensitive content from PDFs with true content removal — by coordinates, text patterns, regex, presets, or saved templates.

`POST /v1/redact`

Remove sensitive content from a PDF. Forme performs true redaction — text operators are removed from the PDF content stream, not just covered with a black box. Metadata is automatically scrubbed on every redaction.

```
Authorization: Bearer forme_sk_...
```

***

## Request Body

| Field        | Type     | Required | Description                                                                           |
| ------------ | -------- | -------- | ------------------------------------------------------------------------------------- |
| `pdf`        | `string` | Yes      | Base64-encoded PDF bytes                                                              |
| `redactions` | `array`  | No       | Coordinate-based redaction regions                                                    |
| `patterns`   | `array`  | No       | Text search patterns (max 50)                                                         |
| `presets`    | `array`  | No       | Built-in pattern presets                                                              |
| `template`   | `string` | No       | Slug of a saved [redaction template](/concepts/redaction-templates) (hosted API only) |

At least one of `redactions`, `patterns`, `presets`, or `template` is required. All can be combined in a single request — patterns from all sources are merged.

***

## Coordinate Redaction

Redact specific rectangular regions by page coordinates.

**RedactionRegion fields:**

| Field    | Type     | Required | Description                               |
| -------- | -------- | -------- | ----------------------------------------- |
| `page`   | `number` | Yes      | Page number (0-indexed)                   |
| `x`      | `number` | Yes      | X coordinate in points from the left edge |
| `y`      | `number` | Yes      | Y coordinate in points from the top edge  |
| `width`  | `number` | Yes      | Width in points                           |
| `height` | `number` | Yes      | Height in points                          |
| `color`  | `string` | No       | Redaction box color (default: black)      |

```bash theme={null}
curl -X POST https://api.formepdf.com/v1/redact \
  -H "Authorization: Bearer forme_sk_abc123..." \
  -H "Content-Type: application/json" \
  -d "{
    \"pdf\": \"$(base64 -w0 document.pdf)\",
    \"redactions\": [
      {\"page\": 0, \"x\": 100, \"y\": 200, \"width\": 150, \"height\": 20}
    ]
  }" \
  --output redacted.pdf
```

***

## Text Pattern Redaction

Find and redact text by literal string or regular expression.

**RedactionPattern fields:**

| Field          | Type     | Required | Description                                                        |
| -------------- | -------- | -------- | ------------------------------------------------------------------ |
| `pattern`      | `string` | Yes      | Text to search for (literal string or regex)                       |
| `pattern_type` | `string` | Yes      | `"Literal"` or `"Regex"`                                           |
| `page`         | `number` | No       | Restrict to a specific page (0-indexed). Omit to search all pages. |
| `color`        | `string` | No       | Redaction box color (default: black)                               |

<CodeGroup>
  ```bash curl theme={null}
  curl -X POST https://api.formepdf.com/v1/redact \
    -H "Authorization: Bearer forme_sk_abc123..." \
    -H "Content-Type: application/json" \
    -d "{
      \"pdf\": \"$(base64 -w0 document.pdf)\",
      \"patterns\": [
        {\"pattern\": \"Jane Doe\", \"pattern_type\": \"Literal\"},
        {\"pattern\": \"\\\\d{3}-\\\\d{2}-\\\\d{4}\", \"pattern_type\": \"Regex\"}
      ]
    }" \
    --output redacted.pdf
  ```

  ```javascript Node.js theme={null}
  import fs from 'fs';

  const pdf = fs.readFileSync('document.pdf').toString('base64');

  const res = await fetch('https://api.formepdf.com/v1/redact', {
    method: 'POST',
    headers: {
      Authorization: 'Bearer forme_sk_abc123...',
      'Content-Type': 'application/json',
    },
    body: JSON.stringify({
      pdf,
      patterns: [
        { pattern: 'Jane Doe', pattern_type: 'Literal' },
        { pattern: '\\d{3}-\\d{2}-\\d{4}', pattern_type: 'Regex' },
      ],
    }),
  });

  const redacted = Buffer.from(await res.arrayBuffer());
  fs.writeFileSync('redacted.pdf', redacted);
  ```

  ```python Python theme={null}
  import requests
  import base64

  with open("document.pdf", "rb") as f:
      pdf_b64 = base64.b64encode(f.read()).decode()

  res = requests.post(
      "https://api.formepdf.com/v1/redact",
      headers={
          "Authorization": "Bearer forme_sk_abc123...",
          "Content-Type": "application/json",
      },
      json={
          "pdf": pdf_b64,
          "patterns": [
              {"pattern": "Jane Doe", "pattern_type": "Literal"},
              {"pattern": r"\d{3}-\d{2}-\d{4}", "pattern_type": "Regex"},
          ],
      },
  )

  with open("redacted.pdf", "wb") as f:
      f.write(res.content)
  ```

  ```go Go theme={null}
  pdf, _ := os.ReadFile("document.pdf")

  body, _ := json.Marshal(map[string]any{
  	"pdf": base64.StdEncoding.EncodeToString(pdf),
  	"patterns": []map[string]string{
  		{"pattern": "Jane Doe", "pattern_type": "Literal"},
  		{"pattern": `\d{3}-\d{2}-\d{4}`, "pattern_type": "Regex"},
  	},
  })

  req, _ := http.NewRequest("POST", "https://api.formepdf.com/v1/redact", bytes.NewReader(body))
  req.Header.Set("Authorization", "Bearer forme_sk_abc123...")
  req.Header.Set("Content-Type", "application/json")

  res, _ := http.DefaultClient.Do(req)
  defer res.Body.Close()

  out, _ := os.Create("redacted.pdf")
  io.Copy(out, res.Body)
  ```
</CodeGroup>

***

## Presets

Built-in patterns for common sensitive data types. Pass an array of preset names:

| Preset          | Matches                                      |
| --------------- | -------------------------------------------- |
| `ssn`           | US Social Security Numbers (XXX-XX-XXXX)     |
| `email`         | Email addresses                              |
| `phone`         | US phone numbers                             |
| `date-of-birth` | Date patterns (MM/DD/YYYY, YYYY-MM-DD, etc.) |
| `credit-card`   | Credit card numbers                          |

```bash theme={null}
curl -X POST https://api.formepdf.com/v1/redact \
  -H "Authorization: Bearer forme_sk_abc123..." \
  -H "Content-Type: application/json" \
  -d "{
    \"pdf\": \"$(base64 -w0 patient-record.pdf)\",
    \"presets\": [\"ssn\", \"email\", \"phone\", \"date-of-birth\"]
  }" \
  --output redacted.pdf
```

***

## Redaction Templates

Reference a saved set of patterns and presets by slug. Create templates in the [dashboard](https://app.formepdf.com/dashboard/redaction-templates) or via the dashboard API.

```json theme={null}
{
  "pdf": "<base64>",
  "template": "hipaa-patient-record"
}
```

Combine a template with additional inline patterns — all sources are merged:

```json theme={null}
{
  "pdf": "<base64>",
  "template": "hipaa-patient-record",
  "patterns": [{"pattern": "Jane Doe", "pattern_type": "Literal"}],
  "presets": ["credit-card"]
}
```

<Note>Redaction templates require the hosted API. Self-hosted users should pass `patterns` and `presets` directly.</Note>

***

## Response

`200 OK` with `Content-Type: application/pdf` — the redacted PDF.

***

## What Gets Removed

* **Text operators** (`Tj`, `TJ`, `'`, `"`) within redacted regions are removed from the PDF content stream
* **Metadata** is automatically scrubbed: author, creator, edit history, and comments are stripped. `/Producer` is replaced with "Forme" and `/ModDate` with the current timestamp.
* **Visual overlay**: A black rectangle covers each redacted area

<Tip>To verify redaction worked, open the PDF in Chrome and try to select text in the redacted area. If the text is truly removed, there will be nothing to select.</Tip>

***

## Known Limitations

* **WinAnsi encoding only**: Text search works on standard Western-encoded text. CJK and other non-WinAnsi text is not searchable by patterns — use coordinate redaction instead.
* **Maximum 50 patterns per request**: Combine multiple presets into a single request, but keep the total pattern count under 50.
* **Images and vector graphics**: Visual black rectangles cover these elements, but the underlying image data is not removed from the content stream in v1.

***

## Related

* [Redaction](/concepts/redaction) — how Forme redaction works, compliance use cases
* [Redaction Templates](/concepts/redaction-templates) — creating and managing saved templates