Installing Claude Code and Your First Launch
Before you can have Claude Code do anything for you, you need to get it installed and running on your machine. This lesson walks through the practical steps: what you need beforehand, how to install the tool, how to sign in, and what to expect the very first time you launch it inside a project. None of this requires deep technical knowledge, but if you have never used a terminal before, take this lesson slowly and do not skip the explanations of what each step is actually doing.
What a Terminal Is, Briefly
If you are completely new to the terminal, here is the short version: it is a text-based way of controlling your computer, where instead of clicking icons you type commands and press enter. On a Mac this is usually an application called Terminal, on Windows it might be the Command Prompt, PowerShell, or a tool like Windows Terminal, and on Linux it is typically just called a terminal or shell. Claude Code lives and runs inside this environment. It can feel intimidating at first, but you only need to know a handful of things to get started, and Claude Code itself will often help you with the rest once it is running.
What You Need Before Installing
Claude Code has a couple of prerequisites worth checking first. You will need Node.js installed on your machine if you plan to install via the npm package manager, since that is the tool used to fetch and install the package. Alternatively, Anthropic also provides a native installer for common operating systems that does not require Node.js to already be present, which can be a simpler path if you are not planning to do general JavaScript or Node development. You will also need an Anthropic account in order to authenticate, since Claude Code needs to verify who you are and connect to Claude models to function.
Installing the Tool
There are two common ways to get Claude Code onto your machine, and which one suits you depends mostly on whether you already have Node.js installed.
| Method | When to use it |
|---|---|
| Install via npm (npm install -g @anthropic-ai/claude-code) | Use this if you already have Node.js and are comfortable running a command in a terminal to install a global package. |
| Native installer | Use this if you would rather download and run a standalone installer for your operating system without setting up Node.js first. |
Once installation finishes, you have a new command available in your terminal called “claude”. Typing this command is how you start every Claude Code session, and it is worth remembering because you will use it constantly from this point forward.
Signing In
The first time you run Claude Code, it will ask you to authenticate. This usually involves logging in with your Anthropic account, and Claude Code may open a browser window as part of this process to confirm your identity before returning control to the terminal. Depending on how your account is set up, you may be authenticating through a standard Anthropic account or through API access; either way, the tool will guide you through the necessary steps the first time, and you generally only need to do this once per machine.
Your First Launch, Step by Step
Here is what a typical first run looks like, broken into a simple sequence. Do not worry if the exact wording on your screen differs slightly, since the tool is updated regularly; what matters is understanding the sequence of events.
- Open your terminal application.
- Navigate to a project folder you want Claude Code to work in. This matters a great deal: Claude Code operates on the folder you launch it from, treating that folder as the project it can see and act on. If you are just experimenting, create a small empty folder for practice rather than launching it somewhere sensitive.
- Type “claude” and press enter.
- Complete the authentication step if this is your first time.
- You should now see a conversational prompt, ready for you to type a message to Claude, much like a chat interface but running inside your terminal.
Why the Folder You Start In Matters
New users are sometimes surprised that Claude Code is scoped to the directory they launched it from. This is a deliberate and important design choice, not a limitation. It means Claude Code’s attention, and its ability to read and change files, is naturally bounded to the project you are actually working on, rather than your entire computer. When you want to work on a different project, you generally start a fresh session from that project’s folder. As you progress through this course, get into the habit of checking which folder you are in before you launch Claude Code, since this single habit prevents a large share of beginner confusion.
A Quiet First Check
Once you are inside a session, a good way to confirm everything is working, before asking Claude to change anything at all, is to ask it a simple, low-stakes question about the project you are in, such as asking it to describe what files are present or what the project appears to do. This does not require Claude to take any risky action, but it does exercise the basic reading and understanding capability you learned about in the previous lesson, and it is a reassuring first step before you move on to asking it to actually make changes, which is where the next lesson picks up.
If Something Goes Wrong
Installation problems are common for first-time terminal users and are rarely a sign that anything is fundamentally broken. The most frequent issues are a missing or outdated Node.js installation, a terminal that has not been restarted after installation so it does not yet recognize the new command, or an interrupted authentication step. If the “claude” command is not recognized right after installing, closing and reopening your terminal window is often enough to fix it, since terminals sometimes need to reload their list of available commands. Patience here pays off, because everything else in this course depends on having a working, authenticated installation to practice with.
