
Free, local-first CLI and macOS menubar app that tracks AI coding token usage and cost across 36+ tools like Claude Code, Cursor, Codex, and Gemini CLI.
Free, local-first CLI and macOS menubar app that tracks AI coding token usage and cost across 36+ tools like Claude Code, Cursor, Codex, and Gemini CLI.
CodeBurn is a local-first observability tool for developers who use multiple AI coding assistants and want to know exactly where their tokens and dollars are going. It reads the session files that tools like Claude Code, Cursor, Codex, Copilot, Gemini CLI, Kiro, OpenCode, and Goose already write to disk, prices every call through LiteLLM, and produces a breakdown by task, model, tool, and project — with no proxy, wrapper, or API key required. A deterministic classifier labels every AI turn as Coding, Debugging, Feature Dev, Refactoring, Testing, and other categories without any LLM inference. A separate `codeburn optimize` command grades the developer's setup A–F, flags duplicate reads, context bloat, and ghost agents, and can apply fixes with per-change undo. CodeBurn is 100% free, open source under a permissive license, and ships as a terminal TUI and a macOS menubar app.
CodeBurn is a free, local-first command-line interface (CLI) and macOS menubar application that efficiently tracks AI coding token usage and costs across more than 36 tools, including Claude Code, Cursor, Codex, and Gemini CLI. It empowers developers to monitor their AI tool expenses seamlessly.
CodeBurn is designed for developers who utilize multiple AI coding tools. It allows users to monitor token usage and associated costs in real-time. By integrating with tools like Claude Code and Codex, users can gain insights into how much they are spending on AI services.
CodeBurn works by consolidating cost data from over 36 AI coding tools into a unified dashboard. It employs a local-first architecture for privacy, classifies tasks deterministically, and provides detailed breakdowns of project and model costs, helping developers and teams optimize their AI usage and budget effectively.
CodeBurn is designed to enhance visibility into AI coding tool expenses by reading on-disk session files from a wide array of more than 36 different AI coding tools, including Claude Code, Cursor, Codex, and Copilot. This integration allows developers to have a centralized dashboard that reflects their AI usage and related costs.
By aggregating data from multiple tools, CodeBurn provides a comprehensive view of token usage and expenses. For instance, an independent developer using both Claude Code and Cursor can quickly identify which tool consumes the most resources, enabling them to adjust their workflows accordingly.
CodeBurn stands out with its local-first architecture, meaning that it does not rely on external proxies, wrappers, or API keys. This ensures that all computations and data processing occur directly on the user's machine, enhancing privacy and security. Since the session files are already generated by the tools, CodeBurn utilizes these files without transmitting any sensitive information over the internet.
CodeBurn intelligently classifies every interaction with the AI tools into distinct categories such as Coding, Debugging, Feature Development, Refactoring, and Testing. This deterministic bucketing process eliminates the need for additional Large Language Model (LLM) calls, making the system efficient and straightforward. For example, if a developer experiences an explosion in token usage, CodeBurn can immediately highlight which task category caused the spike, allowing for quick adjustments.
Developers can view costs and token counts on a per-project and per-model basis. This feature is especially helpful for engineering managers who need to attribute AI costs to specific product areas or clients, ensuring accurate budgeting and resource allocation.
codeburn optimize: This command grades your AI tool setup, identifying issues such as duplicate file reads or excessive context that can inflate costs. Regular optimization helps maintain efficiency.CodeBurn features robust multi-tool cost tracking, a local-first architecture for privacy, task classification for efficiency, detailed project breakdowns, and an optimization grader. These functionalities help developers manage costs and enhance productivity across multiple AI coding tools seamlessly.
CodeBurn offers several standout features designed to empower developers using AI coding tools:
Multi-tool Cost Tracking: CodeBurn reads session files from more than 36 AI coding tools such as Claude Code, Cursor, Codex, Copilot, and others. This unification into a single dashboard enables developers to monitor their usage and costs effectively.
Local-First Architecture: CodeBurn operates without relying on external proxies, wrappers, or API keys. This means all computations occur locally, ensuring that sensitive data never leaves the developer's machine. This architecture is particularly beneficial for organizations concerned about data security and compliance.
Task Classification: The software automatically categorizes AI interactions into specific tasks like Coding, Debugging, Feature Development, Refactoring, and Testing. This deterministic classification helps developers understand how their resources are being utilized, allowing for more informed decision-making without the need for additional LLM calls.
Per-project and Per-model Breakdowns: CodeBurn provides granular insights by slicing costs and token counts by project, model, tool, and task. This feature is crucial for identifying which repositories or models contribute most significantly to expenses, helping developers optimize their workflows.
codeburn optimize Grader: This unique grading feature assesses a developer's setup on an A to F scale, highlighting issues like duplicate file reads, context bloat, and ghost agents. With a single command, developers can apply suggested fixes and even undo changes if necessary, streamlining the optimization process.
codeburn optimize command regularly to maintain a clean and efficient development environment.CodeBurn is designed for indie developers, engineering managers, and teams utilizing AI tools. It provides insights into token consumption, budget allocation, and session performance, allowing users to optimize their workflows and manage AI costs effectively.
CodeBurn offers a suite of features that cater specifically to different user personas in the software development ecosystem:
Solo Developer Cost Visibility: For indie developers, CodeBurn is invaluable in tracking which coding sessions consume the most tokens. For instance, by running Claude Code and Cursor side by side, developers can easily identify sessions that lead to excessive token usage, allowing them to adjust their workflows to be more efficient and cost-effective.
Team AI Budget Attribution: Engineering managers can leverage CodeBurn to roll up AI expenditures by project, making it easier to attribute costs to specific product areas or clients. This feature is crucial for teams working on multiple projects, as it provides a clear breakdown of where AI resources are being utilized and helps in budget forecasting.
Debugging Runaway Sessions: CodeBurn's classification system helps users pinpoint which tasks lead to runaway sessions that consume thousands of tokens. For example, if a coding agent's task unexpectedly escalates in token usage, CodeBurn identifies the specific category that caused the spike, allowing developers to troubleshoot and refine their approach.
Optimizing Agent Setups: By running the command codeburn optimize on a local machine, users can assess their AI setups and eliminate redundant context that may inflate prompt costs. This optimization can lead to significant savings in token usage over time.
Comparing Model Economics: Developers can analyze the economic viability of various AI models using CodeBurn. By examining the per-model breakdown, they can decide whether to transition specific tasks from more expensive frontier models to more cost-effective alternatives, maximizing their budget efficiency.
CodeBurn is entirely free to use, offering users a range of features without any subscription or hidden fees. This makes it an attractive option for individuals and businesses looking to leverage coding tools without financial commitment.
CodeBurn stands out in the marketplace as a free coding tool, making it accessible for students, hobbyists, and professionals alike. Users can perform tasks such as code generation, debugging, and even collaboration on projects without incurring any costs.
For instance, if you're a student learning programming, you can practice coding in different languages like Python, JavaScript, or C++ without worrying about subscription fees. Similarly, small businesses can utilize CodeBurn to develop software solutions without a financial burden, allowing them to allocate resources elsewhere.
Additionally, CodeBurn frequently updates its platform, ensuring users have access to the latest coding features and tools. This commitment to improvement means that even as a free service, users can expect a robust and evolving platform.
By utilizing CodeBurn's free offerings, users can enhance their coding skills and productivity without any financial investment.
To get started with CodeBurn, visit CodeBurn's official website and sign up for an account. Once registered, you can explore its features, create projects, and leverage its powerful tools for efficient coding and development.
To begin using CodeBurn, follow these steps:
By following these guidelines, you will be well on your way to utilizing CodeBurn effectively for your coding projects.
Compare CodeBurn: vs Ito · vs Kin Health · vs Caveman · vs BearDrive