This project provides a self-hosted, unlimited, and free web API for accessing advanced AI models like DeepSeek R1 and GPT-4o. It's designed for users who want to integrate powerful AI capabilities into their applications or workflows without incurring costs or relying on external paid services.

How It Works

The server leverages the g4f library to interface with various AI providers, abstracting away the complexities of direct API calls. It supports multiple models and providers, with a default to Bing and GPT-4. The architecture is built on Flask with asynchronous capabilities (aiohttp), enabling efficient handling of requests. An optional GUI simplifies configuration and usage, while features like proxy support and cookie management aim to ensure continuous access and avoid rate limiting.

Quick Start & Requirements

• Install via pip: pip install -r requirements.txt
• Run server: python3 FreeGPT4_Server.py [--enable-gui] [--port PORT] [--model MODEL] [--provider PROVIDER]
• Requirements: Python 3, Flask[async], g4f, aiohttp, aiohttp_socks, auth, Werkzeug.
• Docker: docker container run -p YOUR_PORT:5500 d0ckmg/free-gpt4-web-api
• Documentation: https://github.com/aledipa/Free-GPT4-WEB-API

Highlighted Details

• Supports self-hosted DeepSeek-R1 and GPT-4o.
• Offers unlimited and free usage.
• Includes a user-friendly GUI for configuration.
• Supports proxy usage to avoid blocking.
• Integrates with Siri via an Apple Shortcut.

Maintenance & Community

The project welcomes contributions. The README mentions credits to specific users for contributions and notes that the demo server may be overloaded.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README. This may pose a restriction for commercial use or closed-source linking until clarified.

Limitations & Caveats

The project relies on third-party providers and the g4f library, which may be subject to changes or instability affecting API availability. The "free" nature implies potential rate limits or unreliability from the underlying providers. The lack of a specified license is a significant caveat for adoption.