0%
Get more from your
Claude subscription.
Rate limits, context health, and token usage — always visible in your macOS menu bar.
If you can, please support the project by purchasing on the App Store or sponsoring on GitHub. If not, it's completely free to download and use. macOS 13+ · MIT License
Dashboard
You're paying for Claude.
Are you getting your money's worth?
Pro costs $20/month. Max costs up to $200. But Claude Code gives you almost no visibility into how you're using it — so you can't tell if you're getting the full value of what you're paying for.
Rate limits are invisible
You hit the wall mid-task with no warning. No countdown, no percentage — just "please wait." No way to pace yourself or plan around resets.
Context fills up silently
Sessions fill with stale file reads and old tool results. By the time quality drops, you've burned tokens on degraded responses you'll redo.
No usage breakdown
Millions of tokens flow through your sessions. Without the numbers, you can't tell if your tier is right or where your budget goes.
Maximize every dollar
of your subscription.
One click on the menu bar icon. Everything you need to use Claude Code efficiently, in a native macOS popover.
Rate Limits
See your 5-hour burst and 7-day sustained usage at a glance. Utilization percentage, reset countdowns, and time-to-limit predictions so you can pace your work.
Context Health
Track your 5 most recent sessions with live context fullness. Smart warnings for long conversations, bloated input, and rapid token consumption.
Token Usage
Per-model breakdown: input, output, cache read, and cache write. Toggle API cost estimates to see what your usage would cost at pay-per-token rates.
Activity & Insights
Sparkline charts across 24 hours, 7 days, and 12 months. Today's stats with trend arrows, projections, and your all-time busiest day.
Smart Alerts
Get notified on Anthropic outages and when your usage crosses a threshold you set. Plan around limits instead of being surprised.
Multi-Account
Support for 2 Claude accounts with one-click switching. Perfect for separate personal and work orgs — each with its own limits and stats.
Your context window is a budget.
Spend it wisely.
Claude can hold ~200K tokens in memory per session — think of it as a notepad. Every message you send, every file Claude reads, every response it gives goes on that notepad. Nothing gets erased between turns. Once the notepad is full, Claude Code auto-compacts: it summarizes everything into a few paragraphs and throws the rest away. That summary is lossy — specific file contents, exact instructions, and nuanced decisions get compressed into a brief recap. Claude continues working, but it's now working from a summary instead of the real thing.
Why this matters
- Before auto-compact, a packed context means Claude is scanning through thousands of irrelevant old tokens on every turn — responses get slower and less accurate.
- After auto-compact, Claude loses specifics — it might re-read files it already read, forget constraints you set, or miss details from earlier in the session.
- You need roughly 20–40% free context to give Claude enough room to think well. Below that, you're paying for degraded output.
- Under 60% — plenty of room. Claude has full recall of everything in this session.
- 60–80% — getting crowded. Quality starts to slip — consider running /compact or starting fresh.
- Over 80% — auto-compact is imminent. Start a new session to get the best results.
Download free from GitHub
Open source, MIT-licensed, every feature included. No account required.
brew tap KyleNesium/tap && brew install --cask aibattery
curl -sL https://github.com/KyleNesium/AIBattery/releases/latest/download/AIBattery.zip -o /tmp/AIBattery.zip && ditto -x -k /tmp/AIBattery.zip /Applications && xattr -cr /Applications/AIBattery.app && open /Applications/AIBattery.app
Requires macOS 13+ and Claude Code. If it saves you time, consider sponsoring the project or buying the App Store version when it launches.