Tokimeter is a local usage analytics tool designed for developers and individuals who frequently use various AI coding assistants. It consolidates usage data from multiple AI tools into a single, comprehensive report, allowing users to monitor their token consumption, project costs, and adherence to usage limits. The tool is particularly useful for those who need to manage expenses associated with AI-driven coding and writing.
The proliferation of AI coding tools has led to fragmented usage data, making it difficult for users to track their overall expenditure and understand where their AI credits are being consumed. This lack of centralized analytics can result in unexpected costs and inefficient resource management. Tokimeter addresses this problem by aggregating the usage records that these AI tools already generate locally on a user's machine.
One of Tokimeter's core features is its ability to pull and consolidate usage records from a wide array of AI coding tools. This includes popular options like Claude and Codex, available in both CLI and desktop applications. It also supports advanced IDEs and platforms such as Cursor, Grok Build, Hermes, opencode, Cline, and Copilot CLI. By centralizing this data, users gain a unified view of their AI tool consumption.
The tool provides detailed breakdowns of usage, allowing users to analyze costs and exact token counts across various dimensions. This includes filtering by project, session, day, specific tool, and even the AI model used. This granular level of detail empowers users to identify patterns, optimize their usage, and allocate their AI budgets more effectively.
Tokimeter incorporates helpful features for budget management, such as 5-hour and weekly limit windows. It provides budget warnings directly in the user's status line, alerting them when they are approaching or exceeding their defined limits. This proactive notification system helps prevent overspending and ensures users stay within their allocated budgets.
For users who need to bill clients or track project-specific expenses, Tokimeter offers a reporting feature that generates per-project files. The `--md` flag, for instance, creates a markdown file that can be easily attached to invoices, providing a clear and transparent breakdown of AI tool usage for each client or project.
Tokimeter operates entirely locally on the user's machine, ensuring privacy and security. It requires no account creation and collects no telemetry data, meaning that no information leaves the user's computer. This commitment to privacy is a key aspect of its design, making it a trustworthy solution for sensitive development workflows.
The benefits of using Tokimeter include enhanced cost control over AI tool usage, improved understanding of AI consumption patterns, and increased efficiency in managing AI resources. By providing clear, consolidated analytics, it helps users make informed decisions about their AI tool adoption and expenditure.
Concrete use cases for Tokimeter include individual developers tracking their personal AI coding tool expenses, freelance developers generating detailed reports for client invoices, and teams managing shared AI tool budgets. It's also valuable for users experimenting with multiple AI models and tools to determine which ones offer the best cost-performance ratio for their specific tasks.
Tokimeter is open-source and MIT licensed, with a free tier for its core local analytics features. A 'Pro' version is available for $4/month, offering features like history that survives log pruning and cross-machine synchronization. The tool is primarily a web application accessible via CLI commands and integrates with various AI coding tools. The CLI is available via npm install.
In summary, Tokimeter offers essential local usage analytics for AI coding tools, providing developers with the insights needed to manage costs, track token usage, and optimize their AI workflows securely and privately.