ccx

Why we included this project

CCX sits between your application and the upstream model providers, translating Claude, OpenAI, Codex, and Gemini request formats so a client written for one API can talk to another without a rewrite. The bundled web admin console handles channel management, key rotation, health checks, and failover, which spares you most of the plumbing that multi-provider setups usually demand. Routing is context-aware: it filters candidate channels by each model's actual context window and max output, so requests land on backends that can actually serve them. Everything runs behind a single port with embedded frontend assets, which makes it a practical fit for small teams or personal services that want resilient, multi-key access to several LLM providers without standing up extra infrastructure.