CodeBurn is a cost-tracking tool that monitors AI coding token usage and expenses across multiple AI-assisted development platforms and models.
The tool addresses the problem of opaque spending on AI coding assistants by running locally and aggregating usage data from supported platforms including Claude Code, Cursor, Codex, Gemini, and others. It tracks tokens consumed and associated costs broken down by model, project, and task, giving developers visibility into where their AI coding budget actually goes. The tool operates as a command-line utility accessible via npx and also provides a web dashboard and menubar interface for monitoring.
Developers should adopt this tool if they use multiple AI coding assistants and want to understand their actual spending patterns across these tools. It suits teams and individuals who need cost visibility without relying on vendor-provided billing dashboards, which may not offer the same granularity or cross-platform aggregation. The tool is free and runs locally, meaning usage data stays on the user's machine rather than being sent to external services.
The project shows active development with ongoing work to expand platform coverage and improve observability features. The maintainers actively solicit community feedback about discrepancies between the tool's reported usage and official vendor bills, indicating a commitment to accuracy. The codebase is maintained in TypeScript and distributed through npm, with engagement channels including Discord and social media for user support and feature requests.
