Understand how Claude Code uses tokens across projects, branches, sessions and models. Track estimated costs, cache performance, budgets and optimization opportunities through a private local dashboard with no telemetry or API key required.
Hey Product Hunt 👋
CC Token Meter started with a simple question:
Where are my Claude Code tokens actually going?
Total usage alone wasn't enough. I wanted to understand which projects and sessions were driving consumption, whether caching was helping, what the estimated cost looked like, and where there were opportunities to use Claude Code more efficiently.
That's what CC Token Meter is built around.
It turns the Claude Code session data already stored on your machine into a local dashboard for:
⚡ Live usage
💰 Estimated cost
🧠 Cache intelligence
📊 Project and session attribution
🎯 Budgets
💡 Evidence-backed optimization insights
Privacy was a hard requirement: transcripts remain read-only, there is no telemetry, and the application doesn't require an Anthropic API key.
It's completely open source under the MIT License.
I'm particularly interested in feedback from people using Claude Code daily:
What information about your AI coding usage would help you work differently?