AI Launchpad

Your AI stack, running in minutes.

Install, authenticate, and run the tools shaping agentic development. Pick your operating system, copy the commands, and go.

Copy-paste ready

Every command verified against official docs.

macOS, Linux & Windows

Switch platforms and commands update instantly.

Track your setup

Check off tools and keep progress in your browser.

Before you start: install Node.js 22+ for npm-based tools, keep Git handy, and have an account or API key for the tools you want to use.

Your setup progress0/6 · 0%

Codex CLI

Terminal agent

OpenAI · Included with ChatGPT Plus / Pro

GPT in your terminal

OpenAI's lightweight coding agent. It reads your repository, edits files, runs commands, and works through multi-step engineering tasks from a single prompt.

ChatGPT account or OpenAI API keymacOS, Linux, or Windows

1Install

curl -fsSL https://chatgpt.com/codex/install.sh | sh

Prefer npm? Run: npm install -g @openai/codex

2Authenticate

codex

Choose "Sign in with ChatGPT" on first launch, or configure an API key.

3Start a session

cd my-project && codex

4Try it

codex "Explain this codebase and suggest three improvements"

Docs GitHub Website

Google Antigravity

Agentic IDE

Google · Free with a personal Google account

Gemini agents, IDE + terminal

Google's agent-first development platform. Use the Antigravity app or IDE for a visual workspace, or the agy CLI for a terminal-first multi-agent experience powered by Gemini.

macOS 13+, Windows 10+, or LinuxGoogle account

1Install the CLI

curl -fsSL https://antigravity.google/cli/install.sh | bash

Want the IDE instead? Download it from antigravity.google/download

2Sign in

agy

Complete the browser sign-in with your Google account, then finish the terminal setup.

3Start a session

cd my-project && agy

4Try it

agy -p "Add unit tests for the authentication handlers"

Docs Website

Claude Code

Terminal agent

Anthropic · Requires a paid Claude plan or API access

Agentic coding in your terminal

Anthropic's terminal-native coding agent. It understands entire codebases, edits multiple files, runs tests, handles git workflows, and connects to tools through MCP.

Claude Pro, Max, Team, Enterprise, or Console accountmacOS, Linux, WSL, or Windows

1Install

curl -fsSL https://claude.ai/install.sh | bash

Prefer npm? Run: npm install -g @anthropic-ai/claude-code

2Authenticate

claude

Follow the browser login prompt. An ANTHROPIC_API_KEY can be used instead.

3Start a session

cd my-project && claude

4Try it

claude "Review the last commit and fix any failing tests"

Docs GitHub Website

GitHub Copilot CLI

Terminal agent

GitHub · Requires a GitHub Copilot subscription

Your repo, issues, and PRs from the shell

GitHub's terminal agent with native access to your repositories, issues, and pull requests. Ask it to explain code, implement a fix, or open a PR without leaving the shell.

Node.js 22+ (npm install)Active GitHub Copilot subscription

1Install

npm install -g @github/copilot

macOS/Linux alternative: brew install --cask copilot-cli

2Authenticate

copilot

Run /login inside the session and follow the device-code prompt.

3Start a session

cd my-project && copilot

4Try it

copilot "Explain the architecture of this repository"

Docs GitHub Website

Ollama

Local models

Ollama · Free and open source

Run open models locally

Run Llama, Gemma, Qwen, DeepSeek, and other open models on your own machine. No API keys, no per-token bill, and your data never leaves your computer.

8 GB+ RAM for 7B modelsGPU recommended for larger models

1Install

curl -fsSL https://ollama.com/install.sh | sh

macOS alternative: brew install ollama

2Start the server

ollama serve

The macOS app starts the server automatically. Skip this step if it is already running.

3Pull and run a model

ollama run llama3.1

4Try it

ollama run llama3.1 "Write a Python script that renames files by date"

Docs GitHub Website

Gemini API

API

Google · Free tier available; pay-as-you-go for higher limits

Call Gemini from your own code

Use Gemini directly in scripts, agents, and backends. Install the SDK, export an API key, and make your first model call in under a minute.

Python 3.9+ or Node.js 18+Gemini API key from Google AI Studio

1Install the SDK

pip install -U google-genai

JavaScript/TypeScript alternative: npm install @google/genai

2Set your API key

export GEMINI_API_KEY="your-key-here"

Create a key at aistudio.google.com/apikey

3Make your first call

python -c "from google import genai; print(genai.Client().models.generate_content(model='gemini-flash-latest', contents='Explain MCP in one sentence.').text)"

4Try it

python -c "from google import genai; print(genai.Client().models.generate_content(model='gemini-flash-latest', contents='Write a haiku about AI agents.').text)"

Docs Website

Want to go deeper?

Read the guides or add your own tool to the directory.

Browse guides Submit a tool