MyBurn

📈 Daily Usage

📅 Activity iWeekly Pattern shows average cost per weekday. Hourly Pattern shows which hours you code most (avg $/hour). Contribution Calendar shows daily activity over the past months — darker = higher cost.

🤖 Model Distribution iBreakdown of burn value by model. Opus is the most capable (and expensive), Haiku is used for fast sub-agent tasks, Sonnet sits in between.

—

Total Burn

📁 Project Breakdown iBurn value per project. Projects are detected from Claude Code's working directory. Useful to see which codebase is burning the most.

📊 30-Day Projection iEstimated burn for the next 30 days based on your recent usage patterns. Three modes available: Last Day (aggressive), Last Week (averaged), and Smart (median of last 7 active days × activity rate).

Loading projection...

⚡ Cache Efficiency iHow efficiently Claude caches context between turns. Higher cache read = less re-processing = lower cost per token. Cache read tokens cost 10x less than fresh input tokens.

Cache Read iTokens served from cache — previously processed context reused at $1.50/M instead of $15/M (Opus). The bigger this is relative to Direct Input, the better.

—

Cache Creation iTokens written to cache for future reuse. Costs $3.75/M (Opus) — a one-time investment that pays off when the same context is read back in subsequent turns.

—

Direct Input iFresh input tokens — not cached, processed at full price ($15/M Opus). Includes new messages and any context that couldn't be cached.

—

Direct Output iOutput tokens — Claude's responses. Always billed at full output rate ($75/M Opus). Cannot be cached.

—

☠️ YOLO Mode iYOLO mode (bypassPermissions) lets Claude execute tools without confirmation. Tracks how often you live dangerously.

Loading YOLO data...

🛠 Tool Usage iTool usage breakdown across all Claude Code turns. See which tools Claude uses most and how they group by category.

Loading tool data...

⚡ Turn Insights iTurn insights — response times, thinking usage, error rates, and session metrics from all Claude Code turns.

Loading turn data...

💡 Model Economics iPer-model cost breakdown showing input vs output token costs. Output tokens are ~5x more expensive than input. Cache savings shown where applicable.

Loading economics...

🔥 Sessions iIndividual Claude Code sessions with cost, duration, and token counts. Click column headers to sort. Use the search box to filter by project name.

Project	Last Active	Tokens	Cost ↓	Cache %	Models
Loading sessions...

—

Welcome to MyBurn

Keyboard Shortcuts

⚙️ Profile & Settings

Login with GitHub