CodeBurn

Tracks AI coding costs and token usage for local AI coding sessions

Coding & Development Coding AssistantsPaid

Visit website ↗

FreeStarting price

30KMonthly visits

25Popularity

About CodeBurn

CodeBurn monitors spending and token usage for AI coding agents such as Claude Code, Cursor, and Codex by reading local session files without uploading data. It attributes costs to specific projects, activities, and pull requests, helping users identify wasteful sessions, duplicate reads, and context bloat. The tool provides a live spend tracker in the menu bar, detailed breakdowns by model, project, and activity, and efficiency scores based on one-shot rates and cache hits. It supports over 40 AI coding tools and runs locally on macOS, Windows, and Linux, with no account or configuration required. Users can view spending from today to lifetime, projected monthly totals, and reconcile costs with commits and PRs. The terminal command offers a quick dashboard without modifying existing workflows.

Key features

Live spend tracker in menu bar
Cost attribution to projects and pull requests
Efficiency scoring based on one-shot rates and cache hits
Terminal command for quick access
Multi-platform support (macOS, Windows, Linux)
Detailed breakdowns by model, activity, and session
Projected monthly spending totals
MCP server and subagent cost tracking

Use cases

Monitoring daily AI coding expenses in real time
Identifying wasteful or duplicate AI sessions
Reconciling AI costs with project deliverables

Pros

Local-only processing with no data upload
Supports over 40 AI coding tools
Menu bar and terminal dashboard for real-time tracking
Attribution of costs to projects, PRs, and activities
Open source with no account or configuration needed

Cons

No cloud sync between machines
Limited to supported AI coding tools
No mobile or web interface

CodeBurn videos

Frequently asked questions about CodeBurn

What does CodeBurn do?

CodeBurn tracks spending and token usage for AI coding agents like Claude Code, Cursor, and Codex by reading local session files without uploading any data. It attributes costs to specific projects, activities, and pull requests, helping users identify wasteful sessions and inefficiencies.

Who is CodeBurn for?

CodeBurn is designed for developers and teams who use AI coding tools and want to monitor, understand, and optimize their spending and token usage across multiple agents and projects.

How does CodeBurn track costs without uploading data?

CodeBurn reads local session files generated by AI coding tools, such as transcripts and logs, to calculate spending and token usage. It does not send any data to external servers, ensuring privacy and security.

What AI coding tools does CodeBurn support?

CodeBurn supports over 40 AI coding tools, including Claude Code, Cursor, Codex, Copilot, Gemini, Goose, OpenCode, Kilo Code, Roo Code, and many others. The list continues to expand as new tools are added.

Can CodeBurn help reduce AI coding costs?

Yes, CodeBurn provides efficiency scores based on metrics like one-shot rates and cache hits, and it highlights wasteful sessions, duplicate reads, and context bloat. It also offers fixes to optimize usage and reduce unnecessary spending.

How do I get started with CodeBurn?

You can start using CodeBurn by running the terminal command `npx codeburn`, which launches a dashboard without requiring an account or configuration. It also offers a menu bar app for macOS, Windows, and Linux.