Tracking every Claude tier

You should know where
every token goes.

Tokens burned, dollars spent, model splits — one install, runs in the background.

Leaderboard
npm install -g contextis

Tracks every Claude tier

Opus 4.8Fable 5Sonnet 4.6Haiku 4.5Claude CodeOpus 4.8Fable 5Sonnet 4.6Haiku 4.5Claude CodeOpus 4.8Fable 5Sonnet 4.6Haiku 4.5Claude CodeOpus 4.8Fable 5Sonnet 4.6Haiku 4.5Claude Code

Cost, not just tokens

Know the dollar number, per model.

Opus, Sonnet, Haiku, Fable — each priced at current API rates. The leaderboard ranks by spend by default, so you finally see what your AI habit actually costs.

$1.2k

FAQ

Questions

What is Contextis?+

A usage analytics platform for Claude Code. A CLI reads your local logs, syncs to the cloud, and a dashboard turns opaque AI usage into tokens burned, dollars spent, model splits, streaks, and a global leaderboard.

How long does setup take?+

Under a minute. npm install -g contextis, sign in with GitHub, paste your token. It autostarts in the background from then on.

What does it cost?+

Free. The CLI is on npm and the dashboard is open. No card, no trial.

Where does my data live?+

Your Claude Code logs are read locally and synced to your own account. Only you see your personal stats; the leaderboard shows aggregate token/cost totals.

Which models does it track?+

Every Claude tier — Opus, Sonnet, Haiku, and Fable 5 — with per-model cost based on current API pricing.

Start measuring today.

The longer you wait, the more invisible spend you miss.