Tokens 4 Breakfast helps Mac builders see Claude Code session usage, Claude plan limits, and Anthropic-style token spend in one local workflow instead of mixing three different billing realities into one vague number.
Start free with one Claude workflow — subscription tracking is included at no cost. Upgrade once when you want Claude beside every other AI provider.
Track Claude Code, Claude Pro or Max, and Anthropic API-style token costs from one local Mac menu bar workflow.
Related paths
FAQ
No. This is the broader Claude hub. The Claude Code cost tracker is narrower and focuses on local Claude Code sessions, project usage, and session spend.
Yes for Claude Code usage. The app reads local Claude Code session files on your Mac. Claude Web and provider API features use provider-side access where the provider requires it, while T4B stays local-first.
These pages are intentionally practical: buyer-intent SEO should answer the exact decision a builder is trying to make before asking for a download.
Claude cost is not one number. Claude Code can be tracked from local session files, Claude Pro or Max is a subscription with usage limits, and Anthropic API work is metered by tokens and model choice. A useful Claude cost tracker has to separate those cases instead of pretending a single dashboard number explains everything.
The most useful Claude Code signal is not only today's total. It is which project, model, and session shape created the burn. Tokens 4 Breakfast reads local Claude Code session data and turns it into a menu bar signal so you can see when Opus-heavy work, stale context, or a long agent run is pushing the session off plan.
If you pay a flat Claude subscription, your bill is predictable but your work can still stop at the wrong time. The practical number is remaining runway: the 5-hour session allowance, the weekly allowance, and whether your current pace will hit the limit before reset. That is why this hub links Claude subscription tracking to rate-limit pages, not only cost pages.
API usage changes the question from 'which plan am I on?' to 'which model and token shape am I paying for?' Input tokens, output tokens, cache reads, and reasoning effort all matter. Tokens 4 Breakfast keeps Claude API-style cost awareness next to the rest of the AI stack so Anthropic spend does not live in a separate mental ledger.
Provider dashboards, spreadsheets, and scripts all have a role. The gap is live multi-provider visibility while you are still working.
Best tracked from local session files: project, model, token class, cache behavior, and session shape. That is the developer workflow where an always-visible menu bar signal changes behavior.
Best treated as a runway problem. The subscription cost is fixed, but the 5-hour and weekly limits can still stop a high-value work session at the wrong time.
Best treated as token math. Model choice, output length, cache reads, and reasoning effort drive the bill, so the tracker needs to show the source of the spend.
Workflow bridge
If you want Claude Code detail, start with the Claude Code cost tracker. If you want plan runway, use the Claude Pro usage tracker or rate-limit tracker. If you need token math, use the Anthropic API cost tracker and the token-cost calculator. This page keeps those intents connected without forcing one page to answer all of them badly.
Start with one free providerPrivacy-first analytics
We use optional analytics to understand aggregate website usage and make the product easier to discover. Google Analytics only loads after you accept. No account data, app data, or personal AI usage is sent, and you can change your choice anytime.
Read the privacy policy