API Documentation

Integrate Khmer OCR into your applications. Extract text, classify documents, and translate — all via a simple REST API.

Bearer token auth 60 RPM default API keys via admin

Quick Start

Get your API key from the admin, then make your first OCR request:

curl — OCR a document

curl -X POST https://ocr.angkor-ai.com/v1/ocr \
  -H "Authorization: Bearer sk_prod_your_api_key_here" \
  -F "image=@document.jpg" \
  -F "doc_type=government"

Response

{
  "success": true,
  "document_id": "a1b2c3d4-...",
  "title": "Royal Decree No. 123",
  "category": "government",
  "raw_text": "ព្រះរាជក្រឹត្យ...",
  "structured": { "reference_number": "123/2026", ... },
  "stats": { "chars": 2450, "lines": 45, "processing_ms": 11200, "engine": "angkor-ocr" },
  "credits": 100
}

Authentication

All API requests require a Bearer token. API keys are prefixed with sk_prod_ and issued by the admin.

Request header

Authorization: Bearer sk_prod_a1b2c3d4e5f6...

Important: API keys are shown only once at creation time. Store them securely. If lost, revoke and create a new one.

Endpoints

OCR

POST/v1/ocr50 credits

Extract text from a Khmer document image. Supports JPEG, PNG, WebP (max 20MB). Returns raw text, structured fields, and auto-classification.

Parameters (multipart/form-data)

image          File    Required  — The document image
doc_type       String  Optional  — government | cadastral | identity | handwritten (default: government)
thumbnail      String  Optional  — Base64 data URL for thumbnail storage
force_category String  Optional  — Override auto-classification
force_title    String  Optional  — Override auto-title
group_id       String  Optional  — Group ID for multi-page PDFs
page_number    Integer Optional  — Page number within group (default: 1)

Translation

POST/v1/translate20 credits

Translate extracted Khmer text to English, French, or Chinese.

Body (JSON)

{
  "text": "ព្រះរាជក្រឹត្យ...",
  "target": "en",           // en | fr | zh
  "document_id": "optional" // saves translation to document
}

Documents

GET/v1/documents

List your processed documents, grouped by multi-page PDF groups.

Query parameters

category  String  Optional — Filter by category

GET/v1/documents/:id

Get full document detail including raw text, structured data, and translations.

GET/v1/documents/group/:groupId

Get all pages of a multi-page document.

DELETE/v1/documents/:id

Delete a document and its thumbnail.

GET/v1/folders

Get category counts for document organization.

GET/v1/user

Get account info: email, credits, total usage.

Rate Limits

API key requests are rate-limited per minute. Limits are set per key and communicated via response headers.

Rate limit headers

X-RateLimit-Limit: 60        # Max requests per minute
X-RateLimit-Remaining: 42    # Requests left in current window
X-RateLimit-Reset: 1712345600 # Unix timestamp when window resets

Default: 60 requests per minute. Contact us for higher limits. Exceeding the limit returns 429 Too Many Requests.

Pricing

Action	Credits	Price/page
OCR (per page)	50	$0.10
Translation (per page)	20	+$0.05

Volume	Price/page
1 - 999 pages	$0.10
1,000 - 4,999	$0.09
5,000 - 9,999	$0.08
10,000+	$0.07

Error Codes

Code	Meaning
400	Bad request — missing or invalid parameters
401	Unauthorized — invalid or missing API key
402	Insufficient credits
403	Forbidden — API key scope does not allow this action
404	Resource not found
413	File too large (max 20MB)
415	Unsupported file type (use JPEG, PNG, or WebP)
429	Rate limit exceeded — check Retry-After header
500	Server error — retry the request

Get API Access

hello@ai-angkor.com Telegram