A real-time terminal monitoring tool for Claude AI token usage with ML-based predictions, cost analysis, and a Rich UI.
Claude Code Usage Monitor is a Python-based terminal application that provides real-time tracking and analytics for Claude AI token consumption. It helps developers manage their API usage by monitoring token burn rates, predicting session limits, and analyzing costs across different Claude subscription plans. The tool features a modular architecture with a rich terminal UI, machine learning-based predictions, and intelligent auto-detection of usage patterns.
Developers and teams using Claude AI's API, particularly those on Claude Pro, Max5, or Max20 subscription plans who need to monitor and optimize their token usage and costs. It's especially useful for developers working on long coding sessions or managing multiple concurrent Claude sessions.
Developers choose this tool for its real-time monitoring with configurable refresh rates, machine learning-based P90 percentile predictions for intelligent limit detection, and comprehensive cost analytics across Claude plans. Its unique selling points include automatic plan switching, WCAG-compliant rich terminal UI with adaptive themes, and the ability to analyze historical usage patterns to provide personalized recommendations.
Real-time Claude Code usage monitor with predictions and warnings
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Uses P90 percentile calculations and machine learning to analyze historical usage from the last 192 hours, providing intelligent session limit detection and burn rate forecasts.
Features WCAG-compliant color schemes, adaptive themes that auto-detect terminal background, and configurable refresh rates up to 20 Hz for a polished, professional display.
Tracks model-specific pricing, cache token calculations, and budget projections across Claude Pro, Max5, and Max20 plans, helping optimize subscription costs.
Automatically switches plans and discovers custom limits by analyzing usage patterns, with features like timezone auto-detection and saved preferences for convenience.
README admits 'externally-managed-environment' errors on modern Linux distros like Ubuntu 23.04+, forcing users to rely on workarounds like uv or virtual environments instead of simple pip installs.
Lacks web or GUI interfaces, making it less suitable for team-wide monitoring or integration into dashboards, and requires terminal access for all users.
Exclusively designed for Claude AI's API; doesn't support other AI services without significant modification, limiting versatility in multi-provider environments.