Auto Route

Auto Route is positioned as a routing-style endpoint within an LLM gateway deployment, designed to sit between applications and upstream language model providers. Rather than being a standalone trained model, it acts as an intelligent dispatcher that selects the most appropriate backend model for each incoming request. This routing approach lets developers keep a single integration point while the gateway transparently handles provider selection, failover when a backend errors, and unified request handling across heterogeneous providers.

The gateway layer behind Auto Route exposes an OpenAI-compatible API surface, which means existing client SDKs and tooling can be pointed at the endpoint without code rewrites. Practical strengths include consolidated cost and token tracking per model and per project, automatic failover to healthy providers, and caching to reduce repeated spend and latency. For teams that want flexibility without vendor lock-in, this routing entry offers a convenient abstraction: one model string in client code, many possible backends chosen for you behind the scenes.