Comments
>log in to commentWe built this API to let developers redact PDFs and images programmatically, using Azure OCR, AI PII detection and irreversible rasterized output. It supports async batch jobs, custom PII categories, and GDPR‑ready EU/Swiss hosting. Let us know what works for you and what could be improved 😊
More AI tools →74 free dev tools →
- AI‑powered detection of names, emails, phones, addresses, IBANs, credit cards
- OCR on every page for scanned PDFs, JPG, PNG
- Batch upload with ZIP download, no page limit
- Studio editor for manual adjustments and highlighting
- REST API for automated redaction at scale
Redact PDF AI is a REST API and MCP server that detects PII in PDFs, scans and images and either deletes it irreversibly or replaces names with consistent [Person 1] tags.
- for
- Developers building back-end services, AI agents, or apps that need automated PDF redaction.
- pricing
- freemium
- license
- MIT
- v0.2.12 months ago
Key features
- AI-powered PII detection — Detects names, emails, IBANs, credit cards and other categories using Azure OCR and AI.
- Irreversible rasterized output — Redacted files are flattened so hidden text cannot be recovered.
- Batch & async processing — Upload multiple PDFs in one job and poll for completion via async endpoints.
- Retention modes — Default ephemeral deletion of originals; optional Studio mode keeps files for review.
- Idempotent retries — Supports X-Idempotency-Key for safe retry of POST requests.
- OpenAPI & LLM-ready docs — Provides OpenAPI spec, Markdown docs and machine-readable llms.txt for agents.
Use cases
- Redact customer contracts in a SaaS onboarding pipeline.
- Automatically cleanse scanned invoices before storing in a data lake.
- Enable AI agents to process PDFs without exposing raw PII.
- Pseudonymize a contract or lease before an AI agent summarizes it, then restore the real names locally with the MCP server.
Redact PDF AI pricing
- Free$0First document up to 5 pages free · No credit card needed
- Starter$50/month1,000 pages/month · PDF & image support · Standard email support
- Business$250/month6,000 pages/month · Priority support · Up to 3 seats
Redact PDF AI vs alternatives
 Redact PDF AI |  Postman |  Insomnia |  Replicate |  Directus | |
|---|---|---|---|---|---|
| Best for | Programmatic GDPR-compliant PDF redaction | API design & testing | API debugging | AI model inference | Content management |
| Pricing | Freemium | Subscription | Free | Subscription | Free |
| DevHunt upvotes | 1 | 0 | 0 | 0 | 0 |
| Launched | Nov 2026 | — | — | — | — |
- Redact PDF AI vs Postman: Postman is an API client for testing, not a PDF redaction service.
- Redact PDF AI vs Insomnia: Insomnia is an API client, not a redaction pipeline.
- Redact PDF AI vs Replicate: Replicate runs AI models via API, while Redact PDF AI focuses on PDF redaction.
- Redact PDF AI vs Directus: Directus provides a headless CMS, not PDF redaction capabilities.
Redact PDF AI FAQ
Is there an API to redact PDFs?+
Yes, Redact PDF AI offers a REST API that redacts PDFs and images programmatically using AI PII detection and OCR.
How do I redact a PDF via the API?+
POST the file to /v1/jobs with your API key and desired PII categories, poll the job until status is ‘redacted’, then download the output.
Does the API support batch processing?+
Yes, a single job can include multiple documents and you can run jobs concurrently for high-volume workloads.
Is the service GDPR-compliant and where is data processed?+
Processing runs on Microsoft Azure in the EU (Ireland) and Switzerland North, encrypted in transit and at rest, and documents are never used to train AI models.
Can files be deleted immediately after processing?+
Yes, using the ‘ephemeral’ retention mode deletes originals right after the job completes; otherwise you can delete jobs via the API.
Edited by the makers, based on redact-pdf.ai · Oct 4, 2026. Details may change; check the official site.
What people say
“Redact PDF AI efficiently locates and permanently eliminates sensitive information from PDFs, scans, and images, including personal names, email addresses, ...”
slashdot.org
