About this project

FreeLLMAPI provides a unified /v1 endpoint to access free tiers from 34 LLM providers and over 600 model endpoints. It allows users to stack multiple free quotas to increase total available inference capacity while managing rate limits and provider failures automatically. Key capabilities include: - Unified API: Supports OpenAI-compatible surfaces for chat, completions, embeddings, images, video, and speech, as well as Anthropic Messages and Gemini native surfaces. - Smart Routing: Features six routing strategies with automatic failover and key rotation when providers are rate-limited or return errors. - Model Management: Includes a self-updating model catalog and a React-based admin dashboard for managing encrypted API keys and tracking usage analytics. - Advanced Features: Offers 'Fusion' (multi-model synthesis), prompt compression, sticky sessions for context handoff, and an MCP server for agent introspection. - Deployment: Available as a Docker container, a native desktop app for macOS and Windows, and runs on Node 20+ environments including ARM SBCs. - Integration: Provides automated setup tools for various coding agents and CLIs such as Claude Code, Aider, Cline, and Cursor.