AugustRush/AGPT35

ChatGPT gpt-3.5-turbo API for Free (as a Reverse Proxy)

Welcome to the ChatGPT API Free Reverse Proxy, offering free self-hosted API access to ChatGPT (gpt-3.5-turbo) with OpenAI's familiar structure, so no code changes are needed.

Quick Links

• Join our Discord Community for support and questions.
  • ⚡Note: Your Discord account must be at least 7 days old to be able join our Discord community.

Table of Contents

• Features
• Option 1: Installing/Self-Hosting Guide (Without using any API key)
  • Method 1: Using Docker or Run it with a Chat Web UI using docker-compose
  • Method 2: Your PC/Server (manually)
  • Method 3: Termux on Android Phones
• Option 2: Accessing Our Hosted API (Free)
• Usage Examples
• License

Features

• Streaming Response: The API supports streaming response, so you can get the response as soon as it's available.
• API Endpoint Compatibility: Full alignment with official OpenAI API endpoints, ensuring hassle-free integration with existing OpenAI libraries.
• Complimentary Access: No charges for API usage, making advanced AI accessible to everyone even without an API key.

Installing/Self-Hosting Guide

Using Docker

1. Ensure Docker is installed by referring to the Docker Installation Docs.
2. Run the following command:

   docker run -dp 3040:3040 pawanosman/chatgpt:latest

3. Done! You can now connect to your local server's API at:

   http://localhost:3040/v1/chat/completions
   

   Note that the base URL is http://localhost:3040/v1.

Install with chat web interfaces

✅ You can run third-party chat web interfaces, such as BetterChatGPT and LobeChat, with this API using Docker Compose. Click here for the installation guide.

Your PC/Server

To install and run the ChatGPT API Reverse Proxy on your PC/Server by following these steps:

Note: This option is not available to all countries yet. if you are from a country that is not supported, you can use a U.S. VPN or use our hosted API.

1. Ensure NodeJs (v19+) is installed: Download NodeJs
2. Clone this repository:

   git clone https://github.com/PawanOsman/ChatGPT.git

3. Open start.bat (Windows) or start.sh (Linux with bash start.sh command) to install dependencies and launch the server.
4. Done, you can connect to your local server's API at:

   http://localhost:3040/v1/chat/completions
   

   Note that the base url will be http://localhost:3040/v1

To include installation instructions for Termux on Android devices, you can add the following section right after the instructions for Linux in the Installing/Self-Hosting Guide:

Termux on Android Phones

To install and run the ChatGPT API Reverse Proxy on Android using Termux, follow these steps:

1. Install Termux from the Play Store.

2. Update Termux packages:

   apt update

3. Upgrade Termux packages:

   apt upgrade

4. Install git, Node.js, and npm:

   apt install -y git nodejs

5. Clone the repository:

   git clone https://github.com/PawanOsman/ChatGPT.git

6. Navigate to the cloned directory:

   cd ChatGPT

7. Start the server with:

   bash start.sh

8. Your local server will now be running and accessible at:

   http://localhost:3040/v1/chat/completions
   

   Note that the base url will be http://localhost:3040/v1

   You can now use this address to connect to your self-hosted ChatGPT API Reverse Proxy from Android applications/websites that support reverse proxy configurations, on the same device.

Accessing Our Hosted API

Utilize our pre-hosted ChatGPT-like API for free by:

1. Joining our Discord server.
2. Obtaining an API key from the #Bot channel with the /key command.
3. Incorporating the API key into your requests to:

   https://api.pawan.krd/v1/chat/completions
   

Usage Examples

Leverage the same integration code as OpenAI's official libraries by simply adjusting the API key and base URL in your requests. For self-hosted setups, ensure to switch the base URL to your local server's address as mentioned above.

Example Usage with OpenAI Libraries

Python Example

import openai

openai.api_key = 'anything'
openai.base_url = "http://localhost:3040/v1/"

completion = openai.chat.completions.create(
    model="gpt-3.5-turbo",
    messages=[
        {"role": "user", "content": "How do I list all files in a directory using Python?"},
    ],
)

print(completion.choices[0].message.content)

Node.js Example

import OpenAI from 'openai';

const openai = new OpenAI({
	apiKey: "anything",
	baseURL: "http://localhost:3040/v1",
});

const chatCompletion = await openai.chat.completions.create({
  messages: [{ role: 'user', content: 'Say this is a test' }],
  model: 'gpt-3.5-turbo',
});

console.log(chatCompletion.choices[0].message.content);

License

This project is under the AGPL-3.0 License. Refer to the LICENSE file for detailed information.