CodexBar is a macOS menu bar application designed to monitor the usage limits of various AI providers, including OpenAI Codex and Anthropic Claude. It allows users to view session and weekly usage, along with reset timers, without requiring manual logins to each service. The application utilizes a combination of methods such as browser cookies, local storage, and CLI integrations to track usage data.
CodexBar distinguishes itself through its comprehensive support for a wide range of AI providers, its privacy-focused design prioritizing on-device parsing, and its flexible configuration options. The ability to combine multiple providers into a single status item with an overview tab enhances usability. Furthermore, the bundled CLI and WidgetKit widget provide convenient access to usage information.
- Multi-Provider Support: Tracks usage for numerous AI models including Codex, Claude, Gemini, and more.
- Real-time Monitoring: Displays session and weekly usage with countdown timers and reset indicators.
- Cost & Usage Scan: Performs a local cost-usage scan for Codex and Claude over the last 30 days.
- CLI Integration: Includes a bundled CLI (
[codexbar](/milisp/codexia)) for scripting and CI/CD workflows (Linux builds also available). - Privacy-Focused: Prioritizes on-device parsing and uses browser cookies only when explicitly enabled.
- Customizable Refresh: Offers presets for refresh cadence (manual, 1m, 2m, 5m, 15m).
- WidgetKit Support: Mirrors the menu card snapshot in the macOS WidgetKit.
CodexBar is an actively maintained project with regular updates and a responsive developer. The project has a significant number of stars and forks, indicating community interest. Recent commits and issue activity suggest ongoing development and bug fixes. Comprehensive documentation is available, covering setup, usage, and provider integration.
CodexBar benefits developers, researchers, and power users who rely on multiple AI models. It provides a convenient and transparent way to track their usage, manage costs, and stay informed about service limits. This eliminates the need to individually log into each platform and manually check usage statistics, streamlining workflow and improving resource management.