Engineered for macOS Sequoia

Never hit an invisible
AI rate limit again.

A minimal Apple Dynamic Island for your Mac desktop that tracks rolling token capacity and reset countdowns across Claude, Gemini, Grok, Hermes, and OpenAI down to the exact second.

Buy Lifetime License — $28 Try Interactive Island
Pay Once, Own Forever
100% Client-Side Privacy
Apple Silicon M1/M2/M3/M4 Native
macOS Sequoia Desktop — AetherHUD Dynamic Island
LIVE PREVIEW
✨ Hover over any model logo below to inspect live quota countdowns:
74%
Claude 3.7 Sonnet Reasoning
Rolling 5-Hour Cap 94.2M tokens
28%
Gemini 2.5 Flash 4.0M TPM
Daily Request Limit 3.36M / 4.0M TPM
64%
xAI Grok 3 2h Sliding
Query Quota Window 64 / 100 queries
82%
Hermes 3 (405B) Nous Local
Inference Throughput 84.2 tok/s
72%
OpenAI o3-mini Tier 5
Weekly Reasoning Limit 144 / 200 reqs
RETINA CLARITY

Crafted for macOS Sequoia & Apple Silicon

Native Cocoa frameless window with multi-layer vibrancy blur. Zero pixelation, zero flicker.

AetherHUD 3D Liquid Glass Dynamic Island Animated Demo
AetherHUD on macOS Sequoia Desktop
ENGINEERING & CRAFT

Built to Apple Human Interface Guidelines

Purpose, Craft, and Delight in every sub-pixel interaction.

OPTICAL LIQUID GLASS

True native translucent materials.

Engineered with backdrop-filter: blur(28px) saturate(1.4) and Apple Sequoia specular rim highlights. Blurs and refracts whatever desktop wallpaper sits behind it.

AetherHUD // Dynamic Island 🟢 100% HEALTHY
LOCAL & API TELEMETRY

Zero Keys Needed & Transparent Loopback

Passive local client scraping (~/.claude, Cursor, Aider) plus a built-in local loopback proxy (127.0.0.1:11435) intercepts rate-limit headers in real time with zero latency penalty.

$ export OPENAI_BASE_URL="http://127.0.0.1:11435/v1"
[PROXY] 127.0.0.1:11435 ACTIVE
[HEADERS] x-ratelimit-remaining: 750k
[CLIENTS] Claude Code & Cursor Linked
GLOBAL HOTKEY

Cmd + Shift + Q

Summon or dismiss the Dynamic Island from anywhere in macOS. Pin it above all apps while coding, or let it dock silently on your wallpaper.

⌘ ⇧ Q
CUSTOMIZATION DRAWER

Tailor your HUD down to each model.

Toggle individual model logos ON or OFF, switch between Capacity % Badges vs. Minimal Logo-only mode, and adjust opacity. All settings save instantly in local storage.

Claude 3.7 Sonnet ON
Google Gemini 2.5 ON
xAI Grok 3 ON
OpenAI o3-mini ON
FRONTIER COVERAGE

Every Model Cadence Tracked

No two providers reset rate limits the same way. AetherHUD standardizes them into one unified HUD.

Model Provider Supported Frontier Models Reset Mechanism HUD Telemetry
Anthropic Claude 3.7 Sonnet, Claude 3.5 Haiku, Opus 5-Hour Rolling Token Capacity Sub-Second Countdown
Google Gemini 2.5 Flash, Gemini 2.5 Pro Midnight UTC (05:30 AM IST) Daily RPM Daily Quota Meter
xAI Grok 3, Grok 3 Reasoning 2-Hour Sliding Query Window Query Capacity Ring
Nous Research Llama 3.3 (70B), DeepSeek-R1, Qwen 2.5 System Compute / Local Memory Inference tok/sec
OpenAI o3-mini, o1, GPT-4o Weekly Reasoning Token Caps Weekly Horizon Clock
LIFETIME OWNERSHIP

Pay $28 once. Track forever.

No recurring monthly fees. No subscription fatigue. Own the utility outright with free lifetime updates.

$ 28 one-time payment
Universal macOS DMG Installer (Apple Silicon M1/M2/M3/M4)
7 Frontier Models Supported (Claude, Gemini, Grok, Hermes, OpenAI)
Sub-Second Countdown Resets down to the exact second
100% Client-Side Privacy ($0 telemetry cloud servers)
Global Hotkey (Cmd+Shift+Q) and Desktop Pinning
Instant Automated License Key delivered to your email
Get AetherHUD Pro — $28

Instant download • Apple Pay / Credit Card • 14-day money-back guarantee

COMMON QUESTIONS

Frequently Asked Questions

How does AetherHUD calculate rate limits without cloud servers?

AetherHUD reads directly from your Mac's local developer token caches and status files (such as ~/.claude/stats-cache.json and Ollama process telemetry). It parses rolling headers and calculates the exact reset timers entirely on your machine. Your API keys and code never leave your Mac.

Does it run on Apple Silicon Macs?

Yes. AetherHUD is compiled as a native universal binary for Apple Silicon (M1, M2, M3, M4) running macOS Monterey, Ventura, Sonoma, and macOS Sequoia.

Can I use one license key on multiple Macs?

Yes! Every license key supports activation across up to 3 of your personal Macs (e.g., your MacBook Pro and your Mac Studio desktop).

What if I don't use all 5 models?

You can click the ⚙️ Settings gear in the HUD to open the customization drawer and toggle off any model you don't use. Your HUD dynamically resizes to stay minimal and clean.