Your AI agent.
One command away.

One-command setup for a fully optimized AI agent environment. Local-first with smart model routing. Free tier included.

curl -fsSL https://goosestack.com/install.sh | sh
Get Started Free → Go Pro — $10/mo →

Requires macOS 13+ · Free & open-source · MIT License

Features

What It Does

Everything you need for a production-grade AI agent, configured automatically for your hardware.

📦

One-Command Install

curl -fsSL https://goosestack.com/install.sh | sh — detects hardware, installs everything, configures optimally. From zero to running agent in minutes.

🗣️

Smart Agent Setup

Interactive hatching flow — your agent asks questions and personalizes itself. Choose persona, preferences, and integrations in minutes.

🧠

Token Router (Pro)

Analyzes request complexity and routes to the optimal model. Simple questions use cheap models, complex analysis uses premium ones.

🏠

Local-First

Runs on your Mac. Your data stays on your machine. Local models via Ollama — no API costs, no external dependencies.

🌐

Multi-Language Support

Installer supports English and Russian (more coming). Plus Telegram integration — chat with your agent from anywhere.

🚀

24/7 Always-On

LaunchAgent auto-start, built-in watchdog that restarts on failure, and sleep prevention. Your agent never goes down.

Simple

How It Works

Three steps. Five minutes. That's it.

1

Run the Command

Paste the install command in your terminal. The installer detects your hardware and handles everything.

2

Answer a Few Questions

Choose your agent persona, configure integrations, and set your preferences. Takes about a minute.

3

Agent is Live

Your dashboard opens at localhost:3000. Your AI agent is running, learning, and ready to help.

Credits

Top Up. Start Building.

The GooseStack API provides smart routing that automatically picks the right model for each task — use expensive models only when needed. Buy prepaid token credits and start using cloud models instantly — no API key management needed.

$10 $25 $50 $100

Works on Free & Pro tiers · Prepaid balance — you control spending · No surprises

Pricing

Start Free. Go Pro When Ready.

Both tiers include the GooseStack API with prepaid token credits. Pro unlocks BYOK, smart routing, and more.

Free
$0
Free forever · Open source
  • One-command installer
  • Local models (Ollama) — $0 forever
  • BYOK (bring your own API key)
  • Personalized agent setup (BOOTSTRAP.md hatching)
  • Dashboard access
  • No credit card needed
How it works: Run local models completely free, or bring your own API keys for cloud models. No server dependency, everything runs on your Mac.
Get Started →
Recommended
Pro
$10/mo
Everything in Free, plus:
  • Everything in Free
  • Token Router (smart model routing)
  • GooseStack API proxy (prepaid credits, no key management)
  • Priority support
  • Premium skills (coming soon)
Token Router: Complexity-based model routing — save money without sacrificing quality. API Proxy: Prepaid credits with no API key management.
Go Pro →
FAQ

Frequently Asked Questions

GooseStack is a one-command macOS installer that sets up a fully optimized AI agent environment. It detects your hardware, installs local models, configures memory and embeddings, sets up a web dashboard, and gets your agent running — all in about five minutes.
Yes! The Free tier includes the installer, BYOK, and local models with no server dependency. The Pro tier ($10/mo) adds Token Router, API proxy, and dashboard access.
GooseStack supports all Ollama-compatible local models (Llama, Mistral, Phi, Gemma, and more), plus cloud models from Anthropic (Claude), OpenAI (GPT-4o, o1), Google (Gemini), and others — through the GooseStack API (prepaid credits) or via BYOK on the Pro tier.
BYOK is available on both Free and Pro tiers — bring your own API keys for direct access to cloud providers. Pro tier also includes the GooseStack API proxy with prepaid credits for easier key management.
Yes. GooseStack runs entirely on your Mac. Your conversations, memory, and embeddings stay on your machine. When using cloud models, data is sent only to the model provider you choose — GooseStack never stores or reads your conversations.
The GooseStack API is a managed proxy service that gives you access to cloud models (Claude, GPT-4, etc.) without managing your own API keys. It analyzes request complexity and routes to the optimal model. Simple questions use cheap models, complex analysis uses premium ones. Available on Pro tier only.
System

Requirements

GooseStack runs on any modern Mac. Apple Silicon gets the best performance.

💻
macOS 13+
Ventura or newer
🧠
8 GB+ RAM
16 GB+ recommended
⚡
Apple Silicon
Recommended · Intel OK
💾
15 GB Disk
For models & deps

🪿 Dashboard