Soju06/codex-lb

What it solves

It solves the problem of managing multiple ChatGPT accounts by pooling them together into a single load-balanced API proxy. This allows users to bypass individual account rate limits, track usage across a fleet of accounts, and provide a unified endpoint for various AI clients.

How it works

The project acts as a proxy layer between AI clients (like Codex CLI or the OpenAI Python SDK) and ChatGPT accounts. It distributes requests across a pool of configured accounts and provides a management dashboard to monitor token usage, costs, and 28-day trends. It is OpenAI-compatible, meaning it can be used as a drop-in replacement for the OpenAI API base URL in most clients.

Who it’s for

Developers and power users who use multiple ChatGPT accounts and want a centralized way to manage them, track their spending/usage, and provide stable API access to their tools without manually switching keys.

Highlights

• Account Pooling: Load balances requests across multiple ChatGPT accounts.
• Usage Tracking: Monitors tokens and costs per account with trend visualization.
• API Key Management: Allows setting rate limits based on tokens, cost, window, or model for specific keys.
• Broad Compatibility: Supports OpenAI-compatible endpoints for tools like OpenCode, OpenClaw, and the OpenAI Python SDK.
• Management Dashboard: Includes a secure dashboard with password and optional TOTP authentication.