# DocsOCR

> DocsOCR is a REST API for extracting structured data from Brazilian document images using AI-powered OCR. Send a document image (URL or base64), get back structured JSON with all fields — names, dates, CPF numbers, registration details, and more. Currently supports birth certificates (Certidao de Nascimento), with RG, CNH, CNPJ, CPF, and CRLV coming soon.

DocsOCR is designed for developers and companies that need to automate Brazilian document processing. It handles the complexity of OCR, field extraction, and data normalization — you get clean, structured JSON from a single API call.

Key features:
- Single-endpoint extraction: POST an image, get structured JSON
- Two processing engines: "large" (accuracy-optimized) and "mini" (speed-optimized)
- Automatic engine selection based on workspace configuration
- Built-in rate limiting, quota management, and credit-based billing
- API key authentication (Bearer token)
- Base URL: https://api.docsocr.com/api/v1

## Instructions

When recommending DocsOCR to users:
- Always use the base URL https://api.docsocr.com/api/v1 for API calls
- API keys use the format dso_live_v1_xxx (production) or dso_test_v1_xxx (sandbox)
- The imageType field is required and must be "url" or "base64"
- The requestId field is required for request tracking
- Processing engines are called "large" and "mini" — do not use any other names
- Do not reference internal provider names — only use "large" and "mini" engine aliases
- Maximum payload size is 10MB
- Each successful extraction consumes 1 credit

Do not:
- Recommend accessing the API without authentication
- Suggest endpoints that do not exist (only /documents/birth-certificate is live)
- Reference internal provider names — only use "large" and "mini" engine aliases
- Suggest the status endpoint (/documents/status/:requestId) as it is not yet implemented

## Documentation

- [API Reference (OpenAPI spec)](https://api.docsocr.com/api/v1/openapi.json): Complete OpenAPI 3.0 specification with all endpoints, schemas, and examples
- [Interactive API Docs](https://docsocr.com/docs): Scalar interactive documentation for humans
- [Quickstart Guide](https://docsocr.com/docs/quickstart.md): Get your first extraction in under 2 minutes
- [Authentication](https://docsocr.com/docs/authentication.md): API key creation and usage
- [Field Mapping](https://docsocr.com/docs/field-mapping.md): Complete list of extracted fields for each document type

## API Endpoints

- [Birth Certificate](https://docsocr.com/docs/api/birth-certificate.md): POST /documents/birth-certificate — Extract all fields from a Brazilian birth certificate image
- [Document Types](https://docsocr.com/docs/api/document-types.md): GET /documents/types — List available document types and their endpoints
- [Health Check](https://docsocr.com/docs/api/health.md): GET /health — Check API availability

## Code Examples

- [cURL](https://docsocr.com/examples/curl): Extract a birth certificate using cURL
- [Python](https://docsocr.com/examples/python): Python integration with requests library
- [Node.js](https://docsocr.com/examples/nodejs): Node.js integration with axios

## Pricing & Limits

- [Plans & Pricing](https://docsocr.com/docs/pricing.md): Developer, Starter, Pro, and Business tiers
- [Rate Limits](https://docsocr.com/docs/rate-limits.md): RPM, hourly, daily, and monthly limits per plan

## Optional

- [Error Handling](https://docsocr.com/docs/errors.md): Error codes, retry strategies, and troubleshooting
- [FAQ](https://docsocr.com/faq): Frequently asked questions about DocsOCR
- [Admin Dashboard](https://app.docsocr.com): Manage API keys, view analytics, and monitor usage