Chat with URL API

Welcome to the Chat with URL API! This API uses various tools and libraries to fetch content from a URL, process it, and generate responses using OpenAI's language models. It supports both GET and POST requests.

Table of Contents

• Getting Started
• Environment Variables
• Endpoints
  • GET /api
  • POST /api
• Request Parameters
• Example Usage
• Error Handling
• Dependencies
• License

Getting Started

To get started with this API, follow these steps:

1. Clone the repository:

   git clone https://github.com/RaheesAhmed/chat-with-url.git
   cd chat-with-url

2. Install the dependencies:

   npm install

3. Set up the environment variables:

   Create a .env file in the root of your project and add your OpenAI API key:

   OPENAI_API_KEY=your_openai_api_key

4. Run the development server:

   npm run dev

5. Open your browser:

   Go to http://localhost:3000 to see the API in action.

Environment Variables

• OPENAI_API_KEY: Your OpenAI API key. This is required for accessing OpenAI's models.

Endpoints

GET /api

Returns a welcome message.

Response

{
  "response": "Welcome to Chat with URL...."
}

POST /api

Processes the content from a given URL and generates a response using OpenAI's language models.

Request Parameters query (string): The question or query you want to ask based on the URL content. url (string): The URL of the content you want to process. model (string, optional): The OpenAI model to use (default is gpt-3.5-turbo). temperature (number, optional): The temperature for the OpenAI model (default is 0.7). maxTokens (number, optional): The maximum number of tokens for the OpenAI model (default is 1000). chunkSize (number, optional): The chunk size for splitting the content (default is 1000). chunkOverlap (number, optional): The chunk overlap for splitting the content (default is 200).

Response

The response will contain the generated answer based on the input query and URL content.

{
  "result": "Your generated answer based on the query and URL content."
}

Example Usage

curl -X POST http://localhost:3000/api -H "Content-Type: application/json" -d '{
  "query": "What is the main topic of the article?",
  "url": "https://example.com/article",
  "model": "gpt-3.5-turbo",
  "temperature": 0.7,
  "maxTokens": 1000,
  "chunkSize": 1000,
  "chunkOverlap": 200,
}'

Response

{
  "result": "The main topic of the article is about..."
}

Error Handling

Errors are logged to the console, and the API returns a JSON response with the error message and a status code of 500.

Dependencies

next: The React framework for building server-side rendered applications. @langchain/core: Core utilities for building language chain applications. @langchain/community: Community utilities for building language chain applications. cheerio: Fast, flexible, and lean implementation of core jQuery designed specifically for the server. dotenv: Loads environment variables from a .env file. openai: OpenAI API client. langchain: Language chain utilities and components.

License

This project is licensed under the MIT License.License.md