Show HN: Toktrack – 40x faster AI token tracker, rewritten from Node.js to Rust (github.com)

A new tool called Toktrack has been introduced, designed to track token usage and costs across various AI coding CLIs—Claude Code, Codex CLI, and Gemini CLI—within a unified dashboard. This tool, reprogrammed from Node.js to Rust, significantly enhances performance, allowing users to inspect data that previously took over 40 seconds to process in under one second. By leveraging Rust's speed, along with advanced libraries such as simd-json for ultra-fast JSON parsing and Rayon for parallel processing, Toktrack demonstrates a remarkable throughput of approximately 3 GiB/s. The significance of Toktrack lies in its ability to efficiently manage and visualize token consumption in a multi-CLI environment, making it easier for AI developers to understand their spending on APIs and models. The tool includes a terminal-based user interface (TUI) with four different views for detailed breakdowns—daily, weekly, monthly, and overall stats. Additionally, Toktrack preserves data even when original CLI session files are deleted, with cached daily summaries ensuring usage history remains intact. This functionality, along with its ease of installation across multiple platforms, marks Toktrack as a valuable resource in the AI/ML community for better cost management and analytics.