new-api

Why we included this project

Anyone who has wired up more than one LLM provider knows the pain of rewriting client code every time a vendor changes its API. This project is a single gateway that accepts requests in one format and translates them to whichever upstream provider you point it at, so your application code stays stable while the models behind it rotate. It also covers the operational side that often gets neglected: per-user API keys, token grouping, usage tracking, and cost accounting across all your providers. That makes it a practical fit for small teams or individuals who want one dashboard and one billing view instead of logging into five separate consoles. If you self-host your own model endpoints, the gateway can sit in front of those too, giving you a consistent interface for both external and internal models.