llm-document-ocr - GitLite

KO | EN

gitlite — search

#react #ai #typescript #machine-learning #linux #nodejs #macos #kubernetes #vue #security #pytorch #mcp #openai #open-source

llm-document-ocr

(★ 110)

LLM Based OCR and Document Parsing for Node.js

파일 탐색기

최종 버전 다운로드 (.zip)

- cropImage.ts
- index.ts
- pdf-text-reader.ts
.gitignore
CODE_OF_CONDUCT.md
CONTRIBUTING.md
Dockerfile
LICENSE.md
package-lock.json
package.json
README.md
tsconfig.json
tsup.config.ts

# 설치 가이드

1. 코드 내려받기

git clone https://github.com/mercoa-finance/llm-document-ocr

깃허브에서 프로젝트 코드 전체를 내 컴퓨터로 내려받습니다.

cd llm-document-ocr

방금 내려받은 프로젝트 폴더 안으로 이동합니다.

2. Docker

쉬움 추천

사전 준비물

Git GitHub에서 프로젝트 코드를 내려받으려면 필요합니다.
Docker Desktop 컨테이너를 빌드하고 실행하려면 필요합니다. 설치 후 실행해서 백그라운드에 켜두세요.

docker build -t llm-document-ocr .

Dockerfile을 기반으로 실행 가능한 이미지를 빌드합니다.

docker run -p 8080:80 llm-document-ocr

빌드된 이미지를 실제 컨테이너로 실행합니다.

✅ 터미널에 docker compose ps 를 입력해 컨테이너들이 Up 상태인지 확인하세요. README에 포트 번호가 적혀있다면 브라우저에서 http://localhost:포트번호 로 접속해보세요.

3. Node.js

쉬움

사전 준비물

Git GitHub에서 프로젝트 코드를 내려받으려면 필요합니다.
Node.js npm 명령어를 쓰려면 Node.js가 설치되어 있어야 합니다. LTS(장기지원) 버전을 추천합니다.

npm i --save llm-document-ocr

이 명령어를 터미널에 그대로 입력해 실행하세요.

yarn add llm-document-ocr

이 명령어를 터미널에 그대로 입력해 실행하세요.

✅ 명령어 실행 후 터미널에 나타나는 주소(보통 http://localhost:3000 형태)를 브라우저에서 열어보세요.

이 레포의 README에 적힌 실제 명령어를 그대로 가져왔습니다.

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

<h1 align="center"> <code>llm-document-ocr</code> </h1>

<div align="center">

[![npm version](https://img.shields.io/npm/v/llm-document-ocr.svg)](https://npmjs.org/package/llm-document-ocr "View this project on NPM")
[![npm downloads](https://img.shields.io/npm/dm/llm-document-ocr)](https://www.npmjs.com/package/llm-document-ocr)
[![license](https://img.shields.io/npm/l/llm-document-ocr)](https://raw.githubusercontent.com/mercoa-finance/llm-document-ocr/main/LICENSE.md)

</div>

---

Sponsored by [Mercoa](https://mercoa.com), the API for BillPay and Invoicing. Everything you need to launch accounts payable in your product with a single API!

---

LLM Based OCR and Document Parsing for Node.js. Uses GPT4 and Claude3 for OCR and data extraction.

- Converts PDFs (including multi page PDFs) into PNGs for use with GPT4
- Automatically crops white-space to create smaller inputs
- Cleans up JSON string returned by the LLM and converts it to an JSON object
- Custom prompt support for capturing any data you need

Supports:

- ✅ PNG
- ✅ WEBP
- ✅ JPEG / JPG
- ✅ GIF
- ✅ PDF
- ✅ Multi-page PDF
- ❌ DOC
- ❌ DOCX

## Installation

```bash
npm i --save llm-document-ocr
```

```bash
yarn add llm-document-ocr
```

Note: If you are deploying via Docker, see the [Dockerfile](https://github.com/mercoa-finance/llm-document-ocr/blob/main/Dockerfile) for an example Alpine base image.

## Usage

```ts
import { DocumentOcr, prompts } from "llm-document-ocr";

const documentOcr = new DocumentOcr({
  apiKey: 'YOUR-OpenAi/Anthropic-API-KEY' // required, defaults to process.env.OPENAI_API_KEY. OpenAI models need an OpenAI API key, Antrhopic models need an Anthropic API key.
  model: "gpt-4o", // optional, defaults to "gpt-4-turbo". Options are: "gpt-4-turbo", "gpt-4o", "claude-3-opus-20240229", "claude-3-sonnet-20240229", "claude-3-haiku-20240307"
  standardFontDataUrl: "https://unpkg.com/pdfjs-dist@3.2.146/standard_fonts/", // optional, defaults to "https://unpkg.com/pdfjs-dist@3.2.146/standard_fonts/". You can use the systems fonts or the fonts under ./node_modules/pdfjs-dist/standard_fonts/ as well.
});

const documentData = await documentOcr.process({
  model: "gpt-4o", // optional, defaults to model defined in constructor
  document: 'JVBERi0xLjMNCiXi48/TDQoNCjEgMCBvYmoNCjw8DQ...', // Base64 String, Base64 URI, or Buffer
  mimeType: 'application/pdf', // mime-type of the document or image
  prompt: 'invoiceStartDate, invoiceEndDate, amount', // system prompt for data extraction. See examples below.
  pageOptions: 'FIRST_AND_LAST' // optional, defaults to 'ALL'. Determines which page of the PDF will be processed. Available options are 'ALL', 'FIRST_AND_LAST', 'FIRST', 'LAST'.
})
```

## Prompts

Prompts will be automatically prefixed to tell the LLM to return JSON. You will need to specify the data you wish to extract, and the LLM will return a JSON object with those keys.

For example, the prompt we use at Mercoa for invoice processing is the following:

```js
`invoice number, invoice amount, currency (as ISO 4217 code), dueDate, invoiceDate, serviceStartDate, serviceEndDate,
  vendor's [name, email with @, website],
  line items [amnt, price, qty, des, name, cur (as ISO 4217 code)]`;
```

And this returns a JSON object that looks like:

```ts
{
  invoiceNumber?: string | number
  invoiceAmount?: string | number
  currency?: string
  dueDate?: string
  invoiceDate?: string
  serviceStartDate?: string
  serviceEndDate?: string
  vendor: {
    name?: string
    email?: string
    website?: string
  }
  lineItems: Array<{
    des?: string
    qty?: string | number
    price?: string | number
    amnt?: string | number
    name?: string
    cur?: string
  }>
}
```

## Issues and Contributing

If you encounter a bug or want to see something added/changed, please go ahead
and
[open an issue](https://github.com/mercoa-finance/llm-document-ocr/issues/new)

If you wish to contribute to the library, thanks! Please see the [CONTRIBUTING](https://github.com/mercoa-finance/llm-document-ocr/blob/main/CONTRIBUTING.md) guide for more details.

## License

MIT © [Mercoa, Inc](https://mercoa.com/)