🌊 OpenParse

OpenParse provides an API to accurately extract your unstructured data (e.g. PDF, images, charts) into structured format.

🌱 Set up your OpenParser API key

OpenParse is still in private beta. If you are interested in testing our document models, please reach out at info@cambioml.com for a FREE API key.

To set up your API key CAMBIO_API_KEY, you will need to :

1. create a .env file in your root folder;
2. add the following one line to your `.env file:

   CAMBIO_API_KEY=17b************************
   

💻 Installation

conda create -n openparse python=3.10 -y
conda activate openparse
pip3 install open_parser

:bashfile usage

To use OpenParse via curl requests, you can run the following bash command from the root folder of this repository:

bash parse.sh <your apiKey> <file path> <prompt for parse (optional, default="")>

For example, to extract a table from a PDF file, you can run the following command:

bash parse.sh gl**************************************  /path/to/your/file.pdf "Return the table in a JSON format with each box's key and value."

📜 Examples

OpenParse can extract text, numbers and symbols from PDF, images, etc. Check out each notebook below to run OpenParse within 10 lines of code!

Prompt to Extract Key-values into JSON from W2 (PDF)

Do you want to extract key-values from a W2 PDF into JSON format? Check out this notebook (3-min read)!

Extract a Table from an Image into Markdown Format

Are you a financial analyst who need to extract ACCURATE number from a table in an image or a PDF. Check out this notebook (3-min read)!