Get started

How much is your Claude Code usage really worth?

A Pro or Max subscription shows you one flat charge. Claude Code's own logs know exactly how many tokens you used on which model. MyTokenTracker reads those logs and prices them at API rates, so you can see what you are getting for the money.

See it in ten seconds

npx mytokentracker             # last 30 days, by day, agent and model
npx mytokentracker --plan 100  # compare with Max 5x ($20 Pro, $200 Max 20x)

No sign-up and no upload. It works on macOS, Linux and Windows with Node.js 20 or later, and it picks up Codex, GitHub Copilot CLI and other coding agents in the same report.

What you see

Cost by day and model

Input, output and cache tokens for every Opus, Sonnet and Haiku model you ran, priced at current API list prices.

Plan value

Your usage against your plan price as a multiple. Above 1x, the plan is paying for itself.

History past 30 days

Claude Code clears local transcripts after 30 days by default. An account keeps synced daily totals, across all your machines.

Every coding agent

Claude Code, Codex, Copilot CLI, Antigravity, OpenCode, Amp and more, side by side in one report.

Keep it running

  1. 1

    Create an account

    Sign in with GitHub, Google, Microsoft or an email link, and copy your API token.

  2. 2

    Run npx mytokentracker init

    Paste the token at a hidden prompt. The CLI uploads your history and schedules a sync every 30 minutes on macOS and Linux.

  3. 3

    Open the dashboard

    Spend by day, agent and model, a monthly plan-value receipt, budgets and a weekly email.

Full steps, including Windows: install guide. How each number is calculated: methodology.

Frequently asked questions

How much does Claude Code cost per month?

Claude Code is included with Claude Pro at $20 a month and with Max 5x at $100 or Max 20x at $200 a month. What your usage is worth is a different number: run npx mytokentracker to price the last 30 days at API rates and compare.

How do I check my Claude Code usage and cost?

Run npx mytokentracker in a terminal. It reads the usage logs Claude Code keeps on your machine and prints the last 30 days by day, agent and model, priced at API list rates. No account is needed and nothing is uploaded.

Why does my usage history stop at 30 days?

Claude Code deletes local session transcripts after 30 days by default, and any local report can only read what is still on disk. You can raise the cleanupPeriodDays setting, or connect a free MyTokenTracker account so daily totals are synced and kept.

Does MyTokenTracker see my code or conversations?

No. Without an account nothing leaves your machine. With an account, the CLI uploads daily token totals and cost per agent and model. Prompts, responses, code, file names and paths are never uploaded.

How accurate is the cost?

Token counts come straight from Claude Code's own logs, and each model is priced at its current API list price. It is what the same usage would cost on the API, not what Anthropic bills a subscriber.

Is it free?

Yes. The CLI is open source under the MIT license, and the account and dashboard are free. There are no paid tiers.

Related

Is Claude Max worth it? Your usage against Pro, Max 5x and Max 20x. Claude Code usage analytics Trends, budgets and the weekly email. MyTokenTracker vs ccusage What we add on top of the open source parser. Codex cost tracker The same report for OpenAI Codex CLI.