nextbasedev/super-proxy

Super Proxy is a self-hosted gateway that sits between your applications and different AI providers, like OpenAI, Anthropic, Groq, Gemini, and several others. Instead of writing separate code for each provider's API, you point your app at one endpoint, and Super Proxy translates the request to whichever provider you have configured behind the scenes. The project offers both an OpenAI-compatible interface and an Anthropic-compatible interface, so tools already built for those APIs can often switch to Super Proxy with just a change of base URL and API key. It also includes gateway tokens with per-token usage limits, so an operator running the proxy for a team can control who uses how much, track cost, and check on the health of each connected provider through a built-in dashboard. To run it, you need either Node.js 20 or newer, or Docker with Docker Compose. Setup involves cloning the repository, copying an example environment file, generating two secret keys with openssl, and starting the service either directly with npm or inside a Docker container. Once running, you open the dashboard in a browser, sign in with an admin key, create a user, and issue an API token from there. That token is what your applications use to send requests through the gateway. Super Proxy stores its data in a SQLite database and supports streaming responses, meaning replies can come back gradually instead of all at once, useful for chat-style applications. The project is still pre-1.0, so its configuration and interfaces may still change between updates, and the maintainers point to Git tags and GitHub releases as the source of truth for what has actually been published. It is licensed under Apache 2.0, a permissive license that lets you use, modify, and redistribute the code, including for commercial purposes, as long as you keep the required notices in place.