Built on the right data from day one.

WTClaude shows you what Claude is really costing you — built on the right data, and honest about the rest. It starts with Claude Code in your terminal, where we read the billing-grade numbers Anthropic bills you from.

Now with a web dashboard and opt-in multi-device sync — your real Claude Code numbers, everywhere you work.

npx wtclaude setup
For developers →

Free · open-source (MIT) · 60-second setup.

As featured on

WTClaude

For developers

Live

The accuracy reveal: see your real Claude Code costs — the number behind your bill, not a session-log estimate.

Open the developer page →

For the Complete Claude User

Soon

All your Claude — Chat, Cowork, and Code — in one clear view, with a warning before you hit a limit.

Coming soon — get notified →

Billing-grade for Code in your terminal; an honest estimate for Code in the desktop app. Honest, clearly-labeled estimates for everything else — we never show an estimate as if it were exact.

Three models now. One honest way to choose.

Sonnet 5 is the new Claude Code default. Fable 5 is back. Opus 4.8 is still there for the heavy lifting. WTClaude prices all three from your own real usage — so you choose on what it actually costs you, not the rate card.

Sonnet 5, priced right

The new default (in Claude Code v2.1.197 and on Free/Pro). We price it at $2/$10 — auto-stepping to $3/$15 on Aug 31 — off the model’s own reported tokens, so nothing goes untracked. Cheaper per token than Opus, though its tokenizer runs a bit heavier, so we show you the real number, not the sticker.

Fable 5, back with a clock

Included up to 50% of your weekly limit through ~July 12 (extended from July 7), then usage credits at $10/$50. WTClaude reads its billing-grade cost in your terminal and forecasts what your pace would cost in credits after the allowance — a labeled estimate, with a countdown to the cliff.

Compare all three

wtclaude compare-models re-prices your recorded usage across Opus 4.8, Sonnet 5, and Fable 5 — what would this month cost on each? Split by surface (Code billing-grade, Cowork a labeled estimate, Chat excluded). In the CLI, the dashboard, and the companion.

Comparisons re-price your recorded usage, not the same task run on each model — every projection is a labeled estimate. We surface the cost of the choice; we don’t judge which model is “better.”

npx wtclaude setup

then wtclaude compare-models.

The accuracy story

Why Your Claude Code Cost Tracker Disagrees With Your Bill →

See everything WTClaude does.

Billing-grade accuracy, backup & multi-machine sync, a full dashboard, billing-change readiness — the most complete Claude Code tracker.

Explore all features →