neang-mengseang/LLM-Router

LLM Router is a self-hosted, OpenAI-compatible LLM routing engine that exposes a single API endpoint and automatically routes requests across multiple LLM providers with sequential failover. It’s aimed at developers and teams building LLM applications who need resilience when providers rate-limit, error, or go offline. Key features: drop-in OpenAI-compatible API (works with existing SDKs), SSE streaming, support for 14 providers, virtual models that map aliases to ordered provider/model lists, automatic failover on common error codes, circuit breaker with cooldown for unhealthy providers, direct provider endpoints for testing, per-request headers (provider/model/attempts), config-driven setup via config.yaml, Docker support, and no external dependencies.

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.