WallasAPI: Open-source multi-provider AI router with automatic fallback across 12 providers, built solo by Willen Ponce

WallasAPI routes requests to optimal AI providers (Gemini, Groq, GitHub Models, OpenAI, Mistral, Claude via OpenRouter, and 6 others) based on content type, speed, cost, and availability, with automatic fallback if the primary provider fails.

The router accepts any content (text, image, audio, PDF, video), picks the best provider automatically, and returns responses in OpenAI-compatible format with streaming support — existing OpenAI SDK code works unchanged by pointing to http://localhost:8001/v1.

Free access to 50+ state-of-the-art models available via Gemini, Groq, and GitHub Models without payment; optional paid providers include OpenAI, Mistral AI, and NVIDIA NIM for enterprise models.