Codex Switcher is a desktop application for managing multiple accounts across Codex CLI and related services with automatic failover and relay capabilities.
The tool addresses the problem of managing quota limits and account switching when using multiple Codex accounts or API endpoints. It works by monitoring account quotas in real time and automatically switching between accounts, relay endpoints, or fallback services when limits are reached. The application supports several integration points including GLM, MiMo, and StepFun Step Plan services, allowing users to configure relay endpoints and automatic retry strategies that preserve request integrity across service boundaries. It includes a local proxy component and handles protocol conversions between different API standards, such as converting Codex Responses requests to Chat Completions format when routing through Step Plan.
Developers should adopt this tool if they operate multiple Codex accounts or need sophisticated quota management with automatic failover. It suits workflows where quota exhaustion is frequent and manual account switching is impractical. The project is particularly valuable for those integrating multiple LLM providers, as it handles the complexity of routing requests through different endpoints while maintaining compatibility with Codex's native client expectations. The tool distinguishes itself through support for Luna Reserve quotas, which are separate temporary allocations that the application tracks independently to avoid premature account switching, and through native support for multiple Coding Plan providers without requiring manual model alias configuration.
The project shows active maintenance with recent updates addressing protocol compatibility issues and expanding provider support. Development activity includes bug fixes for relay routing behavior, particularly around WebSocket and SSE bridge handling, and protocol-level corrections such as zstd decompression and function tool parameter validation. The codebase demonstrates attention to edge cases in multi-provider scenarios, with fixes targeting specific issues that arise when routing requests through different service implementations.
