General Thai Document OCR

1 ICper page

v2.1.13-20260822 Active Since Aug 2022 Updated 22 Aug 2026 15,400 pages/hour 94.8% accuracy Whitepaper (PDF) POST /v3/store/ocr/document

The General Thai Document OCR API converts printed Thai and English documents into machine-readable text. It accepts images (PNG, JPEG), PDF files, and Microsoft Office documents, and returns plain text, layout-structured JSON, or generated Microsoft Word, PowerPoint, Excel, or searchable PDF files. The production service processes a page of plain text in 0.53 seconds (median) and sustains 4.28 pages per second — more than 15,000 pages per hour. Measured character accuracy is 94.8 percent on an independent public benchmark of scanned Thai government documents; see Accuracy.

Live Demo

Upload a document, or select one of the examples below.

Try Our AI Demo

Login or create a free account to use this AI service demo and explore our powerful APIs.

Get 100 Free Credits (IC) when you sign up!

Offer ends December 31, 2025

Create Free AccountLogin

Example Images (Click to try)

Try Demo

Try Demo

Try Demo

Try Demo

Try Demo

Try Demo

Try Demo

Try Demo

Try Demo

Try Demo

Try Demo

Try Demo

API Key

Log in to use your API keys or Create an account

Choose Endpoint URL

Endpoint URL

cURL Command

Upload Documents

Quick Start

You need an API key from the iApp AI Portal. Send the document as multipart/form-data:

curl -X POST https://api.iapp.co.th/v3/store/ocr/document/ocr \
  -H "apikey: YOUR_API_KEY" \
  -F "file=@document.pdf"

{
    "text": ["ด่วนที่สุด\nที่ มท ๐๔๐๒.๖ ..."],
    "time": 0.53,
    "iapp": { "page": 1, "char": 1376 }
}

Code examples for Python, JavaScript, PHP, Swift, Kotlin, Java, and Dart are in the Technical Reference.

Endpoints and Pricing

All endpoints accept the same input and charge 1 IC per input page. Choose the endpoint by the output you need.

Endpoint	Output	Price
POST /v3/store/ocr/document/ocr	Plain text	1 IC per page
POST /v3/store/ocr/document/layout	JSON with text, bounding boxes, and component types	1 IC per page
POST /v3/store/ocr/document/docx	Download link to a generated Microsoft Word (.docx) file	1 IC per page
POST /v3/store/ocr/document/pptx	Download link to a generated Microsoft PowerPoint (.pptx) file	1 IC per page
POST /v3/store/ocr/document/xlsx	Download link to a generated Microsoft Excel (.xlsx) file	1 IC per page
POST /v3/store/ocr/document/searchablePDF	Download link to a searchable PDF — the original pages with a selectable text layer	1 IC per page

Every endpoint is also available under the legacy /document-ocr prefix, for example /document-ocr/pptx. For on-premise deployment, see Data Security.

Performance

Measured on the production service, August 2026.

Metric	Value
Median processing time (plain text)	0.53 s per page
Sustained throughput	4.28 pages per second (15,400 pages per hour)
Maximum file size	30 MB
Supported input formats	PNG, JPEG, PDF, DOC, DOCX, XLS, XLSX, PPT, PPTX

Accuracy

Accuracy is measured on an independent, publicly available benchmark: 242 scanned pages from 20 Thai Administrative Court judgments, scored against their original source documents.

Metric	Result
Character error rate (CER)	5.2% (94.8% character accuracy)
Word error rate (WER)	7.0% (93.0% word accuracy)
Document layout analysis	96.97% AP

The benchmark counts every difference between the OCR output and the source document as an error, including headers and typesetting differences, so the figures above are conservative. Clean, well-scanned prints score higher. The full methodology, per-document results, and comparison against a leading global cloud OCR service are published in the benchmark whitepaper:

Download the benchmark whitepaper (PDF)

Data Security and Compliance

• The service is GDPR and PDPA compliant.
• Uploaded documents are used only to produce the requested output.
• Generated download links expire after 10 minutes, and generated files are deleted from the server within one hour.
• A fully self-contained on-premise deployment is available, in which no document data leaves your infrastructure. Contact us for details.

Technical Reference

Request

All endpoints accept multipart/form-data with the apikey header.

Field	Required	Description
file	Yes, unless base64 is used	The document: image, PDF, or Microsoft Office file
base64	No	A single image as a Base64 string, as an alternative to file
email	No	If set, the result is emailed to this address instead of returned in the response

Every response carries the header iapp-input-pages with the number of input pages, which equals the number of IC charged.

Responses

/ocr returns one text string per page:

{
    "text": ["ด่วนที่สุด\nที่ มท ๐๔๐๒.๖ มั่นคงและชุมชนพึ่งตนเองได้\nภายในปี ๒๕๖๕ ..."],
    "time": 0.53,
    "iapp": { "page": 1, "char": 1376 }
}

/layout returns, for every page, the detected components with their text, bounding box (in page coordinates), and component type:

{
    "pages": [
        {
            "page": 1,
            "components": [
                {
                    "text": "หนังสือรับรองผลงาน",
                    "bb_left": 282.3, "bb_top": 165.2, "bb_right": 374.6, "bb_bottom": 184.7,
                    "type": "CenTitle"
                },
                {
                    "text": "หนังสือรับรองฉบับนี้ออกให้เพื่อรับรองว่า ...",
                    "bb_left": 120.1, "bb_top": 194.7, "bb_right": 560.1, "bb_bottom": 271.1,
                    "type": "Para1"
                }
            ]
        }
    ],
    "time": 1.84,
    "iapp": { "page": 1, "char": 1673 }
}

/docx, /pptx, /xlsx, and /searchablePDF return a signed download link. The link is valid for 10 minutes:

{
    "path": "https://api.iapp.co.th/document-ocr/files/pZWNyZXQtZXhhbXBsZS10b2tlbg...",
    "time": 1.72,
    "iapp": { "pages": 1, "char": 1229 }
}

Layout Component Types

The type field of each component in the /layout response is one of the following.

Group	Types
Titles and body text	CenTitle, Para1 – Para5, BodyText3, Text, TextBox
Lists	List, ListNum1 – ListNum4, ListCon1 – ListCon4
Indented blocks	Indent1 – Indent5
Tables	Table, TableLike
Page furniture	Header, Footer, Date, NextPage, UpperLeft, UpperRight, TopRight, BottomLeft, BottomRight
Graphics	Figure, Logo, LeftLogo
Signatures	Signature

Code Examples

cURL

curl -X POST https://api.iapp.co.th/v3/store/ocr/document/ocr \
  -H "apikey: YOUR_API_KEY" \
  -F "file=@document.pdf"

Python

import requests

url = "https://api.iapp.co.th/v3/store/ocr/document/ocr"
headers = {"apikey": "YOUR_API_KEY"}
files = {"file": open("document.pdf", "rb")}

response = requests.post(url, headers=headers, files=files)
print(response.json())

JavaScript

const axios = require("axios");
const FormData = require("form-data");
const fs = require("fs");

const data = new FormData();
data.append("file", fs.createReadStream("document.pdf"));

axios.post("https://api.iapp.co.th/v3/store/ocr/document/ocr", data, {
    headers: { apikey: "YOUR_API_KEY", ...data.getHeaders() },
})
    .then((response) => console.log(response.data))
    .catch((error) => console.log(error));

PHP

<?php
$curl = curl_init();

curl_setopt_array($curl, array(
    CURLOPT_URL => 'https://api.iapp.co.th/v3/store/ocr/document/ocr',
    CURLOPT_RETURNTRANSFER => true,
    CURLOPT_POST => true,
    CURLOPT_POSTFIELDS => array(
        'file' => new CURLFILE('document.pdf')
    ),
    CURLOPT_HTTPHEADER => array(
        'apikey: YOUR_API_KEY'
    ),
));

$response = curl_exec($curl);
curl_close($curl);
echo $response;
?>

Swift

import Foundation

let url = URL(string: "https://api.iapp.co.th/v3/store/ocr/document/ocr")!
var request = URLRequest(url: url)
request.httpMethod = "POST"
request.addValue("YOUR_API_KEY", forHTTPHeaderField: "apikey")

let boundary = "Boundary-\(UUID().uuidString)"
request.setValue("multipart/form-data; boundary=\(boundary)", forHTTPHeaderField: "Content-Type")

var body = Data()
body.append("--\(boundary)\r\n".data(using: .utf8)!)
body.append("Content-Disposition: form-data; name=\"file\"; filename=\"document.pdf\"\r\n".data(using: .utf8)!)
body.append("Content-Type: application/pdf\r\n\r\n".data(using: .utf8)!)
if let fileData = try? Data(contentsOf: URL(fileURLWithPath: "document.pdf")) {
    body.append(fileData)
}
body.append("\r\n--\(boundary)--\r\n".data(using: .utf8)!)
request.httpBody = body

URLSession.shared.dataTask(with: request) { data, response, error in
    if let data = data {
        print(String(data: data, encoding: .utf8)!)
    }
}.resume()

Kotlin

import okhttp3.*
import okhttp3.MediaType.Companion.toMediaTypeOrNull
import okhttp3.RequestBody.Companion.asRequestBody
import java.io.File

val client = OkHttpClient()
val file = File("document.pdf")

val requestBody = MultipartBody.Builder()
    .setType(MultipartBody.FORM)
    .addFormDataPart("file", file.name, file.asRequestBody("application/pdf".toMediaTypeOrNull()))
    .build()

val request = Request.Builder()
    .url("https://api.iapp.co.th/v3/store/ocr/document/ocr")
    .addHeader("apikey", "YOUR_API_KEY")
    .post(requestBody)
    .build()

client.newCall(request).execute().use { response ->
    println(response.body?.string())
}

Java

import okhttp3.*;
import java.io.File;

OkHttpClient client = new OkHttpClient();
File file = new File("document.pdf");

RequestBody requestBody = new MultipartBody.Builder()
    .setType(MultipartBody.FORM)
    .addFormDataPart("file", file.getName(),
        RequestBody.create(MediaType.parse("application/pdf"), file))
    .build();

Request request = new Request.Builder()
    .url("https://api.iapp.co.th/v3/store/ocr/document/ocr")
    .addHeader("apikey", "YOUR_API_KEY")
    .post(requestBody)
    .build();

try (Response response = client.newCall(request).execute()) {
    System.out.println(response.body().string());
}

Dart

import 'package:http/http.dart' as http;

var request = http.MultipartRequest(
    'POST',
    Uri.parse('https://api.iapp.co.th/v3/store/ocr/document/ocr'),
);
request.files.add(await http.MultipartFile.fromPath('file', 'document.pdf'));
request.headers.addAll({'apikey': 'YOUR_API_KEY'});

var response = await request.send();
print(await response.stream.bytesToString());

Limitations

• Printed text only; handwriting is not supported.
• Scans at 200 dpi or higher give the best results.
• Files larger than 30 MB are rejected.

Changelog

Version	Date	Changes
v2.1.13-20260822	22 Aug 2026	Upgraded recognition engine: 15% relative reduction in character error rate on the public benchmark; median per-page processing time reduced from 3.33 s to 0.53 s. Generated files are now served directly by the API with signed, expiring links. Fully self-contained on-premise deployment available. PowerPoint (.pptx), Excel (.xlsx), and searchable-PDF outputs are now generally available.
v2.1.12-20251116	16 Nov 2025	Added the iapp-input-pages response header.
v2.0	Aug 2022	Document layout analysis and Microsoft Word output.