StackMap
Subscribe
Explore / ai-token-monitor
soulduse

ai-token-monitor

Menu-bar/tray app for macOS and Windows that reads Claude Code, Codex and OpenCode session logs and shows live token spend — per-model pricing with cache reads, plan-limit bars, webhook alerts.

280 46 TypeScriptupdated 3 days ago
coding
View on GitHubDispute this mapping →
Curator's take

Zero-setup observability for the coding-agent bill: no API keys, no proxy — it prices the session logs your CLIs already write (cache reads included) and puts today's spend next to the clock, with live 5-hour-session and weekly plan bars that ping Discord/Slack/Telegram before you hit the wall. Offline by default; the leaderboard and chat are strictly opt-in. NOT enforcement or routing — it measures, it doesn't reduce; pair it with a gateway or harness caching fixes when the number scares you. macOS/Windows only, and note the repo currently ships no license file.

Mapped by ShipWithAI editors · links verified
README.md

AI Token Monitor

Release License: MIT

한국어 | 日本語 | 简体中文 | 繁體中文 | Türkçe | Italiano

AI Token Monitor — real-time token & cost tracking for AI coding tools, right from your menu bar

AI Token Monitor is a lightweight system tray app for macOS and Windows that answers one question, all day long: how much are my AI coding tools actually costing me? It reads the local session logs that Claude Code, Codex, OpenCode, and GJC already write, prices every token with per-model rates (cache reads included), and puts today's spend right next to your clock — with charts, plan-limit alerts, an opt-in leaderboard, chat, and webhook notifications one click away.

  • Zero setup — no API keys, no proxies. If you've run Claude Code or Codex once, it just works.
  • Spend at a glance — live cost in the menu bar / system tray, full dashboard on click.
  • Stay ahead of limits — live 5-hour session & weekly plan bars, plus Discord / Slack / Telegram alerts before you hit the wall.
  • Private by design — 100% offline by default; leaderboard, chat, and webhooks are strictly opt-in.
Overview Analytics Leaderboard
Today's usage, 7-day chart, weekly/monthly totals Activity graph, 30-day trends, model breakdown Compare usage with other developers

Download

Download Latest Release

Platform File Notes
macOS (Apple Silicon) .dmg Intel Mac support coming soon
Windows .exe installer Windows 10+ (WebView2 required, auto-installed)

How It Works

How AI Token Monitor works — reads local session logs, parses & prices locally, shows cost in the tray and dashboard

  1. Reads local session logs — watches the JSONL files your AI CLIs already write (see Data Sources for exact paths)
  2. Parses & prices locally — a Rust engine deduplicates entries, applies per-model pricing (cache reads included), and re-aggregates the moment a file changes
  3. Shows it everywhere — tray cost ticker, dashboard charts, plan-limit alert bar, and optional webhook notifications

Everything above happens on your machine. The app makes zero network requests unless you opt in to the leaderboard, chat, or webhooks — and even then only aggregated counts are shared, never code or conversation content.

Features

Feature highlights — track & visualize, compete & share, stay under budget, make it yours

Tracking & Visualization

  • Real-time token tracking — parses session JSONL files from Claude Code, Codex, and OpenCode for accurate usage stats
  • Multi-provider support — switch between Claude / Codex / OpenCode sources, with per-provider cost models
  • Multiple config directories — aggregate work + personal accounts by adding several Claude/Codex roots
  • Daily chart — 7/30 day token or cost bar chart with Y-axis labels
  • Activity graph — GitHub-style contribution heatmap with 2D/3D toggle and year navigation
  • Period navigation — browse weekly/monthly totals with < > arrows
  • Model breakdown — Input/Output/Cache ratio

Continue your stack

What teams reach for next — and why each earns a place beside ai-token-monitor. Ranked by curator confidence.