惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

月光博客
月光博客
D
Docker
腾讯CDC
J
Java Code Geeks
大猫的无限游戏
大猫的无限游戏
The Cloudflare Blog
Martin Fowler
Martin Fowler
MongoDB | Blog
MongoDB | Blog
博客园 - Franky
博客园 - 三生石上(FineUI控件)
Recent Announcements
Recent Announcements
F
Fortinet All Blogs
IT之家
IT之家
WordPress大学
WordPress大学
M
MIT News - Artificial intelligence
爱范儿
爱范儿
Microsoft Azure Blog
Microsoft Azure Blog
Vercel News
Vercel News
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
小众软件
小众软件
N
Netflix TechBlog - Medium
T
Tailwind CSS Blog
Engineering at Meta
Engineering at Meta
博客园 - 【当耐特】

Hacker News - Newest: "AI"

AI can't read an investor deck AI as an attorney? Student uses ChatGPT, Gemini to sue UW over alleged racial discrimination Hacking MCP Servers in AI Systems – The Rug Pull: Tool Changes After Approval GitHub - MeepCastana/KubeezCut: Free Web based video editor Can AI judge journalism? A Thiel-backed startup says yes, even if it risks chilling whistleblowers Coming soon: 10 Things That Matter in AI Right Now DARPA built an AI to fact-check enemy weapons claims What explains heterogeneity in AI adoption? When AI Meets Muscle: Context-Aware Electrical Stimulation Promises a New Way to Guide Human Movements - Department of Computer Science AI Changed How We Build. It Did Not Change What Matters. Linux rules on using AI-generated code - Copilot is OK, but humans must take 'full responsibility for the… Meta spins up AI version of Mark Zuckerberg to engage with employees Code Mode: Let Your AI Write Programs, Not Just Call Tools | TanStack Blog GitHub - Delavalom/graft: Go framework for building AI agents. Type-safe tools, multi-provider (OpenAI, Anthropic, Gemini, Bedrock), zero vendor SDKs. India's TCS tops estimates, says new AI models did not dent services demand Gen Z's fading AI hype Strong feeling: we are in a folded AI reality GitHub - machinarii/total-recall-catalog: A reference catalog of latest knowledge retrieval, memory & RAG systems GitHub - mensfeld/code-on-incus: Give each AI agent its own isolated machine with root, Docker, and systemd. Active defense detects and stops threats automatically.. Quantization, LoRA, and the 8% Problem: Benchmarking Local LLMs for Production AI Iran war: We spoke to the man making Lego-style AI videos that experts say are powerful propaganda Powell, Bessent discussed Anthropic's Mythos AI cyber threat with major U.S. banks GitHub - immartian/bellamem: Persistent belief-graph memory for AI agents. Retrieves decisive context by importance — not recency, not RAG, not /compact. recursive-mode: The Repo-Native Operating System for AI Engineering After the attack on Sam Altman's home, will AI CEO's go on the offensive? The biggest advance in AI since the LLM Opus 4.6 vs GPT 5.4 One Prompt Unity World Generation Test “AI polls” are fake polls Client Challenge Can AI be a 'child of God'? Inside Anthropic's meeting with Christian leaders
Tokens 4 Breakfast — AI Token Usage & Spend Tracker for m...
1Kapish · 2026-06-12 · via Hacker News - Newest: "AI"

Tokens 4 Breakfast

Track token usageand manage AI spendacross major AI platforms.

Ship before the limit hits.

Track subscriptions, token usage, and rate-limit pressure from one private Mac menu bar app.

Claude Code, Cursor, OpenAI, Gemini, Copilot, and more — one local view.

NLNo login

NCNo cloud

NTNo tracking

OPOne-time Pro

Built in Germany · Private by design · Built for AI builders

Inside the app

Four screens that make the price obvious.

Your whole AI stack, visible from one menu bar across the four views you actually check.

Tokens 4 Breakfast menu bar popover showing live Claude Web rate limits, a Usage Insights card, AI subscriptions, and the monthly subscription total.

Live command center

One menu bar. Your whole AI bill.

Live Claude Web limits, your AI subscriptions, and a tap-through to Usage Insights all live in one calm popover. You see the real picture before another session quietly gets expensive.

Tokens 4 Breakfast provider settings showing Claude Web, Claude Code, OpenAI, Cursor, GitHub Copilot, Gemini, DeepSeek, and Mistral connections.

Provider coverage

All 8 providers in one source of truth.

Claude Web connects with live rate limits, Claude Code is auto-detected from your local files, and OpenAI, Cursor, Copilot, Gemini, DeepSeek, and Mistral each take one key. No tab safari.

Tokens 4 Breakfast display settings showing menu bar icon styles and subscription total metric selection.

Menu bar control

Put the number that matters in the menu bar.

Show a live provider, all-provider spend, or subscription total. The signal stays visible in the format that matches how you actually work.

Tokens 4 Breakfast notification settings showing morning digest, Claude Code rate-limit alerts, and budget alerts.

Morning digest + alerts

Get the warning before the bill gets weird.

Morning digests, rate-limit thresholds, and budget alerts surface the drift early. You wake up with the number already waiting.

Key features

The features that make T4B different.

Five product decisions make the app useful while you work, not after the damage is done.

Tokens 4 Breakfast Usage Insights showing tokens by project, share of tokens by model, weekly limit usage, cache reuse, and stretch-your-limit guidance.

Usage Insights (Pro). See where every token goes, by project and by model, what is eating your weekly limit, and how to stretch it. Calculated locally; it never reads your prompts.

Ambient rate-limit warning

The menu bar cup shifts from calm to amber to red before Claude Code rate limits your session.

You do not need another dashboard to tell you your runway is collapsing. The warning stays visible while you are still working.

Month-end bill prediction

Catch the $47/month drift before it turns into another quiet invoice.

The point is not the number alone. The point is seeing the spend pattern early enough to stop it while the month is still recoverable.

$47/mo

Common spend caught

Zero-setup onboarding

Claude Code and GitHub Copilot connect from local config. Cursor, OpenAI, Gemini, DeepSeek, Mistral, and OpenRouter take one key.

This is built for builders in the middle of work, not for people who want another configuration project.

Session budget / Focus Mode

Set a spend cap before you start, then get warned before you drift past it.

The point is to stop the accidental runaway agent run while the session is still open in front of you.

AI Subscription Tracking

Track what you pay for Claude Pro, ChatGPT Plus, Cursor, and every other AI subscription in one place.

See your total monthly AI bill — API costs AND subscriptions combined — before it surprises you.

€132

Typical monthly sub total

Pricing

One price. No subscription. Yours forever.

Start free with the live meter. Upgrade once the forecasting, budget guardrails, and cross-provider visibility start saving real money.

Rate Limit Forecasting

Know you'll hit your limit 2 hours before it happens. Not after.

8 Providers, One Dashboard

Claude, OpenAI, Cursor, Copilot and more — aggregated in real time.

Per-Session Budgets

Set a $2 budget for this task. Get alerted at 80%. Never overspend a session.

A clean funnel, not a crippled trial.

Pick one provider, keep the live cost visible, and decide later if you need the forecasting and control layer.

  • 1 provider of your choice
  • Live menu bar spend display
  • Today / 7-day / 30-day history
  • Claude Code 5-hour window timer
  • Multi-currency support

Download Free

No account required · One provider of your choice · Multi-currency ready

Pro

$7.99

One-time, no subscription ever

Less than two cups of coffee. Pay once, use forever.

Best value

Usage Insights

New in 1.0.6

  • See where every token goes, by project and by model
  • What is eating your weekly limit, with cache-reuse and Opus-share analysis
  • Stretch-your-limit guidance, calculated locally; it never reads your prompts
  • Custom date ranges across every view

Unlock All Providers

  • Includes everything in Free, plus every provider in one view
  • All 8 providers: Claude Code, Claude Web, OpenAI, Cursor, Copilot, OpenRouter, DeepSeek, and Mistral
  • Claude Web subscription rate-limit tracking (5-hour window + weekly)
  • GitHub Copilot Free, Individual, Business, and Enterprise plan support

Intelligence & Forecasting

Unique to T4B

  • Velocity-based rate limit forecasting
  • Plan optimization recommendations with upgrade / downgrade signals
  • Per-provider daily & weekly budget alerts

Workflow & Export

  • Focus Mode session budgets with real-time alerts
  • CSV export & weekly summary share
  • Privacy Audit Log — inspect every network call the app makes
  • Sessions tab with per-session cost breakdown
  • AI subscription cost tracker for Claude Pro, ChatGPT Plus, Cursor, Copilot, and more
  • Multi-currency monthly spend total and provider donut chart
  • Priority email support and lifetime updates

Get Pro — $7.99

One-time payment · macOS 14.6 Sonoma or later · Apple Silicon & Intel · 14-day refund guarantee

Built in Germany

100% local — no cloud, no account

Native Swift — not Electron

Free updates included

14-day refund, no questions

During beta, Pro is free. Click "Get Pro — $7.99" in the app after downloading and it activates instantly.

Why people pay

T4B users find $47/month they forgot was running — on day one.

That is the first moment where the app pays for itself: you catch the drift before it becomes another month of background spend.

No more mid-task interruptions

You know exactly how much runway you have before Claude cuts out. No surprises at minute 47 of a long agent run.

No more end-of-month bill panic

The month-end total is predictable from day one. Tokens 4 Breakfast shows your projected spend every morning.

No more context-switching

The number lives in your menu bar. You never have to open a dashboard, check a browser tab, or log in to anything.

Works with

Works with every AI tool you already use

8 providers. One menu bar. Zero tab switching.

Track your stack

Every high-intent path now points to the right tracker.

Searchers do not all start with the brand. Some look for Claude Code limits, Cursor spend, OpenAI API cost, or a local AI budget tracker. This hub gets each one to the right page in one click.

Compare before you choose

Bottom-funnel pages for people already comparing tools.

Comparison

The alternatives solve parts of the problem. This one stays open while you work.

Single-provider trackers show you a number. T4B shows you what that number means — across every AI tool you use.

Know before you're blocked.

T4B tracks your Claude usage velocity every 5 minutes and projects when you'll hit your 5-hour or weekly limit. You get a heads-up with hours to spare — not after the limit cuts your session.

Budget this task, not your whole month.

Set a $1, $2, or $5 budget for the next hour of work. T4B alerts you at 80% and again at 100% in real time. Perfect for keeping API-billed work sessions under control.

One bill. Eight tools. Full picture.

Single-provider trackers give you one number. T4B aggregates Claude, OpenAI, Cursor, Copilot, and four more into a single dashboard with 7-day history and a 30-day run-rate prediction.

Setup

Most people are fully set up in under two minutes.

The setup flow is short on purpose. Download the app, connect what you use, and let the spend signal stay in the menu bar while you work.

01

Download

Click the button. Open the .zip. Drag to Applications.

02

Connect

Pick your provider in Settings. Claude Code and GitHub Copilot connect automatically from your local config — no API key needed. Cursor, OpenAI, Gemini, DeepSeek, Mistral, and OpenRouter connect via API key. Free plan: 1 provider. Pro: all 8.

03

Watch

Your live spend appears in the menu bar. That's it.

Proof

From developers who stopped guessing

These are the moments the app is built for: switching models sooner, catching runaway sessions earlier, and seeing all provider spend in one place.

I didn't realize I was burning $60/month on Claude Opus for tasks Sonnet handles just fine. Found it on day one.

Alex M., indie developer

The Focus Mode session budget is the only thing that's stopped me from accidentally running a runaway agent loop while I grab coffee.

Priya K., full-stack engineer

I track 4 providers: Claude, OpenAI, Cursor, and Copilot. This is the only app that shows them all in one place.

James T., AI tooling builder

Privacy

Your data never leaves your Mac.

Tokens 4 Breakfast does not need a cloud backend, an account, or a telemetry pipeline. Your usage data stays in a local SQLite database on your machine, so the app keeps working even when you are offline.

German-built, local-first software for people who want the spend signal without turning their usage data into someone else's dataset.

Everything stays on your Mac. No account. No cloud. No subscription — ever.

FAQ

Questions buyers ask before they install.

Quick answers on what T4B tracks, how privacy works, and where the free plan stops.

What is Tokens 4 Breakfast?

Tokens 4 Breakfast (T4B) is a native macOS menu bar app that tracks your AI usage, costs, and rate limits across multiple providers in one unified dashboard. Instead of checking Claude's settings page, then OpenAI's billing page, then Cursor's dashboard separately, T4B aggregates everything into a single live view that's always one click away. It also forecasts when you'll hit rate limits, tracks spending trends over time, and lets you set budget caps with alerts so you never get surprised by a bill.

Where is Tokens 4 Breakfast built?

Tokens 4 Breakfast is built in Germany with a local-first privacy model: no account, no telemetry, and usage data stored on your Mac.

What providers are supported?

T4B supports 8 providers: Claude Code, Claude Web (claude.ai subscription tracking), OpenAI, GitHub Copilot, Cursor, OpenRouter, DeepSeek, and Mistral. Claude Code connects automatically by reading local session files with no API key needed. GitHub Copilot connects via your local CLI credentials. Claude Web connects via your browser session cookie to track 5-hour and weekly rate limits in real time. All other providers connect via API key. The free plan includes 1 provider; Pro unlocks all 8.

What is Focus Mode?

Focus Mode lets you set a dollar budget for a specific work session. Hit Start, pick a budget ($1, $2, $5, or custom), and T4B monitors your real-time spend across all connected providers. You'll get a notification at 80% of your budget and again when you hit 100%. It's useful when you want to timebox AI-assisted tasks without accidentally running up a large bill, especially when using API-billed providers like OpenAI or DeepSeek.

Does the app read my conversations or prompts?

No, and this is not a vague assurance. Here is exactly what happens in the code: when T4B reads Claude Code session files, it extracts only the `usage` metadata object (token counts, model name, timestamp, cost). The actual message content, your prompts, and AI responses are never read, parsed, stored, or transmitted. The same applies to every other provider: T4B calls their official usage/billing API endpoints, which return aggregate token counts and costs, never conversation content. There is no technical path in the app to access your chats.

Is there a free plan?

Yes. The free plan lets you connect 1 provider and includes the full menu bar dashboard, 7-day spending chart, and subscription tracker. It's fully functional with no time limit and no account creation required. If you only use Claude Code and want to track your token usage and costs, the free plan covers that completely.

Does it work with Claude Code without an API key?

Yes. T4B reads Claude Code's local session files (JSONL format in `~/.claude/projects/`) directly on your Mac. No API key, no network call, no configuration beyond enabling the provider toggle. If you use Claude Code, T4B will automatically detect and parse your sessions.