This is a reverse proxy and load balancer designed to manage access to ChatGPT and OpenAI APIs across multiple accounts. It pools accounts for load distribution, tracks per-account token usage and costs over time, enforces per-API-key rate limits by token/cost/window/model, and provides a web dashboard for administration with optional TOTP authentication. The proxy exposes OpenAI-compatible endpoints so it works with Codex CLI, OpenCode, and any standard OpenAI client.