DeclassifAI API

Integrate powerful document redaction and entity detection into your applications with our secure, privacy-focused API

v1.0.0
Fast & Accurate

Three AI engines with 86-92% precision for different use cases

Privacy-First

End-to-end encryption with immediate server-side data deletion

Easy Integration

RESTful API with examples in Python, JavaScript, Go, PHP, and more

Getting Started
Learn how to integrate DeclassifAI API into your application

1. Obtain Your API Key

Sign up for an account and generate your 256-bit API key from the dashboard. The API key is sent as a raw-api-key header in all requests.

2. Choose Your Detection Engine

Fast
Microsoft Presidio - 86% precision, fastest processing
Standard
GLiNER/HIDEME - 89% precision, recommended for most use cases
Advanced
Local hosted model - 92% precision, best for contextual detection

3. Make Your First Request

Upload a PDF document to detect sensitive entities. Select an endpoint from the sidebar to see interactive examples and test directly in the playground.

# Example: Detect Norwegian entities with Fast engine
curl -X POST https://api.declassifai.com/detect/fast \
-H "raw-api-key: your-256-bit-api-key" \
-F "file=@document.pdf" \
-F 'requested_entities=["PERSON_P","EMAIL_ADDRESS_P"]' \
-F "threshold=0.8"

4. Handle Responses

All responses are in JSON format. Check the status code and handle errors appropriately. See the Error Reference section for common issues.

Rate Limits

Free tier: 100 requests/hour. Basic/Pro: 1000 requests/hour. Enterprise: Custom limits. Contact sales for higher limits.

Popular Endpoints
Quick access to commonly used API endpoints
Error Reference
Common HTTP status codes and troubleshooting guide