This project provides an open-source proxy API to access OpenAI's ChatGPT API without requiring a personal API key, aiming to democratize AI access for developers and innovators. It serves as a bridge, forwarding requests to OpenAI's API using a shared key and returning responses, thereby removing financial and access barriers.

How It Works

The project acts as a middleware, accepting requests in the same format as the official OpenAI API. It then forwards these requests to OpenAI's servers using an API key managed by the project. Responses are relayed back to the user, abstracting away the need for individual API key management and associated costs. This approach promotes accessibility and allows broader experimentation with ChatGPT.

Quick Start & Requirements

• Usage: Send POST requests to https://chatgpt-api.shn.hk/v1/.
• Example: curl https://chatgpt-api.shn.hk/v1/ -H 'Content-Type: application/json' -d '{"model": "gpt-3.5-turbo", "messages": [{"role": "user", "content": "Hello, how are you?"}]}'
• Documentation: Refer to OpenAI's official API documentation.

Highlighted Details

• Provides free access to the ChatGPT API, removing the need for personal API keys.
• Acts as a proxy, forwarding requests to OpenAI's API.
• Claims to not collect, record, or store user text or identity information.
• The project has inspired other applications like "Free ChatGPT".

Maintenance & Community

The project was archived on March 26, 2023, due to unsustainable financial costs. Prior to archiving, it received significant community sponsorship. The author advocates for lower OpenAI API pricing and removal of regional restrictions.

Licensing & Compatibility

The repository does not explicitly state a license. The project's goal is to promote open access and innovation.

Limitations & Caveats

The project has been archived due to financial unsustainability, indicating the provided public endpoint may no longer be operational or reliable. Users interested in self-hosting require an OpenAI API key and must manage their own rate limiting.