The Infinite AI Pipeline: Setting Up a Unified and Free LLM API

**Free LLM API** is a self-hosted, open-source tool designed to aggregate the official free tiers of various AI providers—such as **Google, Meta, Mistral, and Nvidia**—into a single, **OpenAI-compatible endpoint** running locally on your machine. By stacking these free tiers, users can access hundreds of millions of tokens every month without the need for paid subscriptions like ChatGPT Plus or Claude Pro. Key Features and Functionality Unified API Access:The tool wraps multiple providers behind a single endpoint, meaning you can point any OpenAI-compatible application (like LangChain or Open WebUI) at `localhost` and manage all your models in one place. Smart Routing and Fallback:It features an automatic fallback chain where you can prioritize models (e.g., Gemini 2.5 for capability or Groq for speed). If a provider is unavailable or hits a rate limit, the router silently moves the request to the next provider in line. Administrative Dashboard: A built-in web UI allows you to manage API keys, track usage analytics, and test prompts in a **playground** that shows exactly which provider served a response and how long it took. * **Full Privacy:** Because the project is **self-hosted**, your API keys and data remain on your local machine rather than passing through a third-party relay. ### **Setup and Implementation** To get started, the system requires **Node.js (version 20 or higher) and Git**. The installation process involves: 1. **Cloning the repository** and installing dependencies using `npm install`. 2. **Creating a secure `.env` file** with an encryption key to protect your stored API credentials. 3. Launching the server via `npm run dev`, which opens the local dashboard. 4. Adding API Keys: You can quickly add free keys from providers like Google AI Studio, Groq, GitHub Models, and Hugging Face by pasting them directly into the dashboard.