MeterTab.app
Know how much AI you have left — a macOS menu-bar meter for Claude Code, Codex and Cursor that your agents can query, and that can stop a run.
A startup from Hungary, founded by Peter Dongo.
Know how much AI you have left — a macOS menu-bar meter for Claude Code, Codex and Cursor that your agents can query, and that can stop a run.
A startup from Hungary, founded by Peter Dongo.
This page is designed to help you find out whether MeterTab.app is good and if it is the right choice for you.
Know how much AI you have left.
MeterTab is a macOS menu-bar meter for your AI usage. It shows the tighter of your 5-hour and weekly limits as one percentage with a live reset countdown, across nine providers: Claude Code, Codex, Cursor, the Anthropic and OpenAI APIs, OpenRouter, Gemini, Groq and Together AI.
It reads what your tools already write. Your AI CLIs keep logs on your Mac. MeterTab parses them read-only after a one-time folder grant. Nothing is uploaded; there is no backend to upload to.
Calibrated, not guessed. Anthropic does not publish token limits, so most trackers assume a denominator. Connect your Claude account and MeterTab uses your account's own percentage, then learns your real limit from it so the gauge stays accurate afterwards.
Your agents can query it. A built-in MCP server runs inside the app over stdio, so no port and no key to leak. An agent can call get_budget mid-task and downshift the model, should_continue and stop cleanly, or get_service_status to tell an Anthropic outage apart from your own limit.
And it can say no. MeterTab installs a Claude Code PreToolUse hook that refuses tool calls once your window is spent or a dollar ceiling you set is breached, with per-worktree limits, a loop guard and a snooze. Every other AI budget enforcer is a proxy, and a proxy forces raw API pricing instead of your subscription. A local hook does not touch your auth.
Also: burn velocity, a 90-day heatmap, per-model cost, real billed dollars via your own API key, threshold notifications, Anthropic service status, CSV export, and native iPhone and Apple Watch apps with widgets and a complication, synced over your own iCloud.
On a subscription the dollar figures are API-equivalent (your tokens at pay-as-you-go rates), and labelled as such.
Free tier forever. Pro is $19.99 one-time, never a subscription, covering Mac, iPhone and Watch. No account, no analytics, no backend.
Listed in
Budget circuit-breaker
Claude Code PreToolUse hook denies tool calls when your window is spent or a dollar ceiling is breached. Per-worktree limits, loop guard, snooze.
MCP server
Runs inside the app over stdio — get_budget, should_continue, get_service_status, spend_by. No port, no key to leak.
Providers tracked
9 — Claude Code, Codex, Cursor, Anthropic & OpenAI APIs, OpenRouter, Gemini, Groq, Together AI
Limit calibration
Uses your Claude account's own percentage, then learns your real limit so the gauge stays accurate after you sign out
Cost reporting
Per-model breakdown, 90-day heatmap, burn velocity, CSV export. Real billed dollars with your own API key.
Service status
Watches Anthropic's status page, so an outage isn't mistaken for your own limit
Platforms
macOS 14+, iOS 17+, watchOS 10+ — one Universal Purchase, with widgets and a Watch complication
Privacy
No backend, no account, no analytics. Keys in the Keychain; iCloud sync carries counts only.
Every AI usage tracker shows you a number. MeterTab is the only one that can act on it. It installs a Claude Code PreToolUse hook that refuses tool calls once your window is spent or a dollar ceiling you set is breached, and it exposes your live budget to your agents over MCP so they can downshift or stop cleanly mid-task. Every other way to enforce an AI budget is a proxy - and a proxy forces raw API pricing instead of your subscription, which is many times the cost. A local hook enforces without touching your auth.
Honestly: if you want a number in a terminal, use ccusage - it's free, open source, and covers more agent CLIs than MeterTab does. Choose MeterTab when you want three things it does that the free tools don't. It watches continuously instead of when you remember to check. Its percentage is calibrated against your Claude account's own figure rather than a guessed token limit. And it can stop a runaway agent instead of just describing one after the fact. It's also one $19.99 purchase covering Mac, iPhone and Watch, with no subscription.
Developers on a Claude Pro or Max subscription who use Claude Code, Codex or Cursor heavily enough to hit the limits - the people running long agent sessions who have lost a window mid-refactor, or burned a weekly cap on a Tuesday. Mostly solo developers and small teams paying $20–$200/month for AI tooling, on macOS.
I'm a solo developer in Hungary and I built it for myself, because Claude Code kept dying mid-task with no warning. Anthropic doesn't publish token limits, so I couldn't tell whether I had an hour of headroom or five minutes. The first version was just a gauge in the menu bar. Then I realised the gauge was solving the wrong half of the problem - I could see the number, but my agent couldn't, and it was the agent doing the spending. So I built the MCP server so agents could ask, and then the hook so the budget could actually say no. That's what turned it from a meter into a brake.
Swift and SwiftUI throughout, with Swift Charts for the graphs, WidgetKit for widgets and the Watch complication, CloudKit for iCloud sync, StoreKit 2 for the purchase, ActivityKit for Live Activities, and the Keychain for API keys. Shared logic lives in a local Swift package. The MCP server runs in-process over stdio. There are no third-party SDKs at all - no analytics, no crash reporting, no attribution - and the only package dependency is our own.
We have collected here some useful links to help you find out if MeterTab.app is good.
Check the traffic stats of MeterTab.app on SimilarWeb. The key metrics to look for are: monthly visits, average visit duration, pages per visit, and traffic by country. Moreoever, check the traffic sources. For example "Direct" traffic is a good sign.
Check the "Domain Rating" of MeterTab.app on Ahrefs. The domain rating is a measure of the strength of a website's backlink profile on a scale from 0 to 100. It shows the strength of MeterTab.app's backlink profile compared to the other websites. In most cases a domain rating of 60+ is considered good and 70+ is considered very good.
Check the "Domain Authority" of MeterTab.app on MOZ. A website's domain authority (DA) is a search engine ranking score that predicts how well a website will rank on search engine result pages (SERPs). It is based on a 100-point logarithmic scale, with higher scores corresponding to a greater likelihood of ranking. This is another useful metric to check if a website is good.
The latest comments about MeterTab.app on Reddit. This can help you find out how popualr the product is and what people think about it.
Do you know an article comparing MeterTab.app to other products?
Suggest a link to a post with product alternatives.
Is MeterTab.app good? This is an informative page that will help you find out. Moreover, you can review and discuss MeterTab.app here. The primary details have been verified within the last quarter. So they could be considered up to date. If you think we are missing something, please use the means on this page to comment or suggest changes. All reviews and comments are highly encouranged and appreciated as they help everyone in the community to make an informed choice. Please always be kind and objective when evaluating a product and sharing your opinion.
