OpenAI builds and deploys advanced AI models like GPT-4o for autonomous agents and workflows.
Burn
0Popularity
About Burn
Burn is a menu bar application designed to monitor AI coding expenditures for users of Claude Code and Codex. It provides real-time insights into current burn rates, projected daily costs, and per-model breakdowns without requiring users to open a separate dashboard. The tool calculates cost per shipped pull request by dividing weekly spend by merged PRs, offering a rough but practical measure of whether AI spending translates into delivered work. Burn supports multiple providers and displays combined or provider-specific spend alongside plan cap progress. It reads PR data from GitHub or self-hosted Git instances such as Forgejo, Gitea, or GitHub Enterprise, with authentication tokens stored securely in the system Keychain. Usage logs are processed locally, and pricing is based on each provider’s published rates, ensuring data privacy.
Key features
- Menu bar interface for quick access
- Live burn rate and midnight projection
- Per-model cost breakdown
- Cost per shipped PR calculation
- Multi-provider support (Claude and Codex)
- GitHub and self-hosted Git integration
- Local usage log processing
- Plan cap progress tracking
Use cases
- Monitoring AI coding costs for individual developers
- Tracking per-model spending efficiency
- Evaluating ROI of AI-assisted pull requests
Pros
- Real-time burn rate tracking with minute-level refresh
- Per-model cost breakdown for Opus, Sonnet, Haiku, and cache operations
- Cost per shipped PR calculation for spend efficiency assessment
- Supports multiple Git hosts including self-hosted instances
- Local data processing with no external dashboard required
Cons
- Mac-only application with no Windows or Linux support
- Limited to Claude Code and Codex providers
- No cloud sync or cross-device functionality
Frequently asked questions about Burn
Where does the cost data come from?
Burn reads Claude Code's and Codex's local usage logs directly on your Mac and prices them against each provider's published rates. Checking usage against your plan's weekly or monthly allowance uses your provider account; everything else stays local.
Does it work with self-hosted Git?
Yes. Add any GitHub-compatible host (Forgejo, Gitea, GitHub Enterprise) from Settings alongside github.com. Tokens live in Keychain and never show up in the UI.
What's 'cost per shipped PR'?
Your weekly (or monthly) spend divided by the number of PRs merged in that window. It's rough, but it's an honest read on whether the spend is turning into shipped work.
Is it free?
Yes. Burn is available as a free menu bar application for macOS.
How does Burn calculate the midnight projection?
The midnight projection estimates your total spend by midnight based on your current burn rate, refreshed every minute.
Can I see spend by individual model?
Yes. Burn provides a per-model breakdown showing costs for models like Opus, Sonnet, and Haiku, including cache read/write impacts.
