Install Claude Code (2026): No Node.js Needed

Who this is for: Complete beginners. You've heard about Claude Code but haven't installed it yet, or you're not sure what a "terminal" is. No coding experience required.

What you'll need:

A computer: macOS 13.0+, Windows 10 (version 1809)+, or Linux (such as Ubuntu 20.04+ or Debian 10+)
At least 4 GB of RAM
An internet connection
A paid Claude account (Pro, Max, Team, or Enterprise) or an Anthropic Console account

The free claude.ai plan does not include Claude Code. You can sign up or upgrade at claude.ai.

You do not need Node.js. The recommended installer works without it.

If this feels intimidating, it's not. The terminal looks technical, but you're really just typing short English words and pressing Enter.

This guide takes you from install to first setup in about 10 minutes.

Once you clear this one hurdle, you can build websites, apps, and automations with Claude Code. You just describe what you want in plain language.

Some coding familiarity helps, but Claude handles most of the heavy lifting.

Prefer to skip the terminal? Claude Code also has a Desktop app. The rest of this guide covers the terminal version.

What is a terminal?

A terminal is a text-based way to talk to your computer. Instead of clicking buttons and icons, you type commands and press Enter.

Every computer already has one built in:

macOS: Open Terminal (search "Terminal" in Spotlight, or find it in Applications > Utilities)
Windows: Open Windows Terminal or PowerShell (search "Terminal" in the Start menu)
Linux: Open your distribution's terminal app (usually Ctrl+Alt+T)

When you open a terminal, you'll see a blinking cursor waiting for input. That's it. Nothing scary. It's just a text box for your computer. (You might hear the word "shell". That's just the program inside the terminal that reads your commands. You don't need to think about it.)

On Windows, look at the start of the line. If it begins with PS C:\, you're in PowerShell. If it shows only C:\, you're in Command Prompt (CMD). This matters in the next section.

A terminal is just a text box — you type, it responds

Your first terminal commands

Try these to get comfortable. Type each one and press Enter:

pwd

This prints your current location (folder). You'll see something like /Users/yourname or C:\Users\yourname.

ls

This lists files in your current folder. On Windows PowerShell, ls also works (it's an alias for Get-ChildItem).

mkdir my-project

This creates a new folder called my-project. You just created a folder without clicking anything.

cd my-project

This moves you into that folder. cd stands for "change directory" (directory = folder).

That's the basics. You now know enough to use Claude Code.

Installing Claude Code

Claude Code is a command-line tool. You install it through the terminal, and you use it through the terminal.

The recommended way is the native installer. It needs no Node.js and no other tools. Pick the section for your system.

The install command shows no progress while it downloads. The terminal may look frozen for a while. That is normal. Wait until you get your prompt back.

macOS, Linux & WSL

Paste this into your terminal and press Enter:

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

Windows PowerShell

Your prompt starts with PS C:\. Paste this and press Enter:

irm https://claude.ai/install.ps1 | iex

You do not need to run PowerShell as Administrator.

Windows CMD

Your prompt shows C:\ without the PS. Paste this and press Enter:

curl -fsSL https://claude.ai/install.cmd -o install.cmd && install.cmd && del install.cmd

On native Windows, we recommend installing Git for Windows. It is optional. WSL does not need it.

Check that it worked

When the install finishes, open a new terminal window. Then run:

claude --version

A working install prints a version number. If you see an error instead, check "Common issues" below.

The native install updates itself in the background. You don't need to do anything.

Log in

Run Claude Code for the first time:

claude

Your browser opens a login page. Sign in with your Claude account (the same one you use at claude.ai). Remember, the free claude.ai plan does not include Claude Code. You need a Pro, Max, Team, or Enterprise plan, or a Console account.

Other ways to install

The native installer is the best default. These options also work:

Homebrew: brew install --cask claude-code
WinGet (Windows): winget install Anthropic.ClaudeCode
npm: npm install -g @anthropic-ai/claude-code (needs Node.js 22 or later)

Homebrew and WinGet do not update automatically. To upgrade, run brew upgrade claude-code or winget upgrade Anthropic.ClaudeCode.

Do not use sudo with npm. If npm gives you a permission error, follow the official fix for permission errors.

What does the `claude` command do?

The claude command starts an interactive session where you can talk to Claude directly in your terminal. It's like chatting with Claude on the web, but with superpowers:

Claude can read your files. It sees your code.
Claude can edit your files. It makes changes directly.
Claude can run commands. It executes shell commands for you.
Claude can use agents. These are specialized personas for different tasks.

Basic usage

Start Claude Code in your project folder:

cd my-project
claude

You'll see a prompt where you can type requests:

> Create a simple index.html with a hello world page

Claude reads your folder, writes the file, and tells you what it did. That's it. You're using Claude Code.

Installed? Your next step is a first setup.

Copy the free Startup Dev Team into your project. You instantly get a product manager, developer, tester and UX designer to work with.

Get the free Startup Dev Team

One-shot mode

You can also give Claude a single task without entering the interactive session:

claude -p "What files are in this project?"

The -p flag runs Claude in non-interactive mode. It prints the answer and exits.

Folder structure: where does everything go?

When people talk about "agents," "skills," and "hooks" in Claude Code, they're talking about files inside a .claude/ folder in your project:

my-project/
  .claude/
    agents/      ← agent persona files (.md)
    skills/      ← slash command workflows
    settings.json  ← hooks and configuration
  src/
  package.json
  ...

The .claude/ folder is how you customize Claude Code for your project. You don't need it to get started. Claude Code works fine without it. But when you're ready to level up, that's where the magic lives.

Common issues

"command not found: claude" or "'claude' is not recognized"

This usually means the install folder isn't on your PATH yet. (PATH is the list of places your computer looks for commands.)

First, close the terminal and open a new window. Then run claude --version again.

Still not working? Follow the official fix for "command not found" after installation.

"The token '&&' is not a valid statement separator"

You are in PowerShell, but you ran the CMD command. Use the PowerShell command instead:

irm https://claude.ai/install.ps1 | iex

"'irm' is not recognized as an internal or external command"

You are in CMD, but you ran the PowerShell command. Use the CMD command instead:

curl -fsSL https://claude.ai/install.cmd -o install.cmd && install.cmd && del install.cmd

Check your setup with claude doctor

If something feels off, run this:

claude doctor

It prints a read-only report about your installation and settings. It changes nothing.

For more help, see the official troubleshooting page.

Claude asks for authentication every time

Make sure you completed the login step (run claude and finish the sign-in in your browser). Your session is saved locally, so you only need to do this once.

What's next?

Now that you have Claude Code running, you can start building with it:

Free Starter Template: Get 4 agents and 5 skills, ready to use. Drop the .claude/ folder into your project.
What Are Claude Code Agents?: Learn how to give Claude specialized roles (product manager, developer, reviewer)
Create Your First Skill: Build reusable slash commands in 5 minutes
From Solo to Team: A practical upgrade path from single assistant to multi-agent team