To start Claude Code in the terminal, open a terminal, paste the one-line installer, open a new window and type claude in the folder you want to work in. The first run opens your browser to log in with a paid Claude plan or a Console account. Then type a task in plain English and press Enter. Below: each step for macOS, Windows and Linux, how to approve or stop an action, how to exit and resume a session, and the errors beginners hit most often.
The Terminal Agent for Absolute Beginners course has 6 lessons of 10 to 15 minutes: from the phone chat to an agent that handles files, PDFs, email and calendar. The first 2 lessons are open without sign-up, and signing up gives 7 days of full access, no card needed.
Updated October 31, 2026. Commands, requirements and plans were checked against Anthropic's Claude Code documentation on that date; the sources are linked in each section.
You need a computer with macOS 13 or later, Windows 10 version 1809 or later, or a recent Linux (Ubuntu 20.04+, Debian 10+), at least 4 GB of RAM, an internet connection and a Claude account on a paid plan. Claude Code works with Pro, Max, Ultra, Team or Enterprise, with a Claude Console account (API access with prepaid credits), or through Amazon Bedrock, Google Cloud or Microsoft Foundry. The free claude.ai plan does not include Claude Code. It also has to be available in your country.
No programming experience is needed. Anthropic's own terminal guide is written for people who have never opened a terminal: you type requests in plain English and Claude does the work. If you would rather skip the terminal entirely, the same agent is available as a desktop app, in VS Code and JetBrains and on the web at claude.ai/code. Sources: Claude Code docs: Advanced setup, Terminal guide for new users.
A terminal is a window where you type commands instead of clicking. It looks empty at first: a line of text and a blinking cursor.
| System | Open it · paste a command |
|---|---|
| macOS | Cmd + Space, type Terminal, press Enter · paste with Cmd + V |
| Windows | Win + X, choose Windows PowerShell or Terminal · paste with Ctrl + V or right-click |
| Linux | Ctrl + Alt + T on most distributions · paste with Ctrl + Shift + V |
On Windows, check which program you opened. PowerShell shows PS C:\Users\YourName> at the start of the line; CMD shows the same path without PS. The two need different install commands. If you see Windows PowerShell (x86) in the title, close it and open the entry without x86: Claude Code does not run in the 32-bit version.
Copy the line for your system, paste it into the terminal and press Enter. The installer shows no progress while it downloads, so leave the window open until you see Claude Code successfully installed!
# macOS, Linux, WSL
curl -fsSL https://claude.ai/install.sh | bash
# Windows PowerShell
irm https://claude.ai/install.ps1 | iex
In Windows CMD (no PS at the start of the line) use this command instead:
curl -fsSL https://claude.ai/install.cmd -o install.cmd && install.cmd && del install.cmd
Homebrew users can run brew install --cask claude-code, and WinGet users winget install Anthropic.ClaudeCode; those two do not update themselves, while the native installer updates in the background.
When it finishes, open a new terminal window and check the install:
claude --version
A working install prints a version number followed by (Claude Code). On native Windows, Anthropic recommends installing Git for Windows as well, so Claude Code can use Bash; without it, Claude Code uses PowerShell. Source: Claude Code docs: Quickstart.
Claude Code works inside the folder you start it from: it reads and changes files there. For a first session, make an empty practice folder so nothing important is in reach, then move into it with cd (change directory):
mkdir ~/claude-practice
cd ~/claude-practice
The same two lines work in PowerShell. Later, to work on real files, cd into that folder instead, for example cd ~/Documents/reports. On a Mac you can type cd with a space and drag the folder from Finder into the window to paste its path.
claude
On the first run Claude Code asks you to log in. A browser window opens; sign in with your Claude account and return to the terminal. Your credentials are stored, so you will not log in again on this computer. When the session is ready, the prompt shows the version, the current model and the folder above it. To switch accounts later, type /login inside the session.
If the terminal answers command not found: claude or 'claude' is not recognized, see the fixes below.
Type what you want in plain English and press Enter. Good first tasks are small and easy to check:
This is my first session. Tell me what you can do on this computer and suggest 3 simple tasks.make me a simple webpage that says hello world (Anthropic's own first example; double-click the HTML file it creates to open it in your browser)what does this project do? once you start Claude in a folder with real filesWhether Claude asks before it changes files or runs a command depends on the permission mode, shown in the status bar. Recent versions (2.1.283 and later) start in auto mode: Claude acts without asking while a classifier reviews each action in the background. For a first session where you want to approve every step, press Shift+Tab once to switch to Manual mode; further presses move to auto-accepting file edits and to plan mode, where edits stay blocked until you approve a plan. When Claude asks, press Enter to choose Yes, or pick another option with the arrow keys; you cannot click inside the terminal. Press Esc to interrupt Claude while it is working. When you are unsure, ask for the plan first: Show me the plan first, do nothing yet. Source: Claude Code docs: Permission modes.
Four rules from the beginner course keep the first weeks safe: verify important facts, read before anything is sent, keep payments and signatures in your own hands, and ask for a plan when in doubt. They are covered in the lesson What to trust and what to double-check.
Type /exit or press Ctrl+D twice on an empty prompt to leave Claude Code. To pick up where you left off, open a terminal in the same folder and run claude -c. claude -c looks for the most recent conversation in the current folder, so run it from the folder you worked in.
| Command | What it does |
|---|---|
claude | Starts an interactive session |
claude "task" | Starts a session with the first request filled in |
claude -c | Continues the most recent conversation in this folder |
claude -r | Lets you pick an older conversation to resume |
claude -p "query" | Answers once and exits, no session |
/resume | Opens a previous conversation |
/clear | Clears the conversation and starts over |
/help | Lists the available commands |
/exit, Ctrl+D twice | Ends the session |
Type / on its own to see every command and skill available to you, and press the up arrow to bring back an earlier request.
command not found: claude (Mac, Linux) or 'claude' is not recognized (Windows). The install folder is not on your PATH yet. Open a new terminal window first. If it still fails, run the command the installer printed under Setup notes; on a Mac with the default Zsh shell that is echo 'export PATH="$HOME/.local/bin:$PATH"' >> ~/.zshrc followed by source ~/.zshrc.'irm' is not recognized. You are in CMD. Open PowerShell with Win + X, or use the CMD install command above.The token '&&' is not a valid statement separator. The reverse: you pasted the CMD command into PowerShell. Use the irm line.syntax error near unexpected token '<'. The install address returned a web page. If it says "App unavailable in region", Claude Code is not offered in your country; otherwise run the command again or install with Homebrew.Every fix above comes from the troubleshooting section of the Claude Code terminal guide.
Claude Code is Anthropic's agent that runs on your computer. Chat in the Claude app writes text back; Claude Code also reads the files in your folder, runs commands and edits files after you approve. People search for it as "Claude agent", "Claude terminal" or "clode code": all of them lead to the same claude command. How an agent differs from a phone chat is explained in plain words in the free lesson What a terminal agent is, and the lesson Launch the agent in 10 seconds repeats this first session as a guided task.
If you write code and want install options, permission modes and your first session on a real repository, continue with the developer lesson Install Claude Code and start your first session.
Every Academl lesson has a short task, a quiz question and an AI tutor; after all lessons of a course you get a certificate with a verification code.
| Course and length | What you learn |
|---|---|
| Terminal Agent for Absolute Beginners 6 lessons of 10–15 min, about 76 min | Agent vs chat, first launch, how to ask, files, PDFs, email, calendar, web search, what to double-check. No coding |
| Claude Code & Agentic Development 10 lessons, about 4 hours | For developers: codebase reading, CLAUDE.md, slash commands, git, debugging, MCP servers, IDE integration, CI |
| AI Coding: Claude Code, Codex, Cursor, Windsurf 4 lessons of 15 min, about 1 hour | Which coding agent fits which job, workflow patterns, rolling it out in a team |
The beginner course is the only course of the From Zero: Terminal Agent for Beginners track.
Time. At about 76 minutes, the beginner course fits into one evening or a week of short sessions; progress is saved between lessons.
Price. The first 2 lessons of every course are free without sign-up; sign-up gives 7 days of full access to the whole catalog with no card, then Pro is $9 a month or $79 a year (pricing). Your Claude plan is paid separately to Anthropic. Questions about access and certificates are answered on the FAQ page.
Open a terminal, go to the folder you want to work in with cd, type claude and press Enter. On the first run a browser window opens for you to log in; after that the prompt appears and you type your task in plain English. To start with a task already filled in, run claude followed by the task in quotes.
Paste the one-line installer from Anthropic into a terminal: the curl command ending in install.sh on macOS, Linux and WSL, or the irm command ending in install.ps1 in Windows PowerShell (both are shown in step 2 above). Open a new terminal window, check claude --version, then run claude and log in with your Claude account in the browser. Credentials are stored, so you log in only once.
No. Anthropic's terminal guide is written for people who have never used a terminal: you describe what you want in plain English and Claude writes the code or does the task. When Claude asks for permission, read the request, then press Enter to approve or Esc to stop.
Claude Code needs a Pro, Max, Ultra, Team or Enterprise plan, a Claude Console account with API billing, or a cloud provider such as Amazon Bedrock. The free claude.ai plan does not include Claude Code. The installer itself costs nothing, and the first two lessons of Academl's beginner course are free.
Type /exit, or press Ctrl+D twice on an empty prompt. Later, open a terminal in the same folder and run claude -c to continue the most recent conversation there, or claude -r to pick an older one from a list. Inside a session, /resume does the same.
The install folder is not on your PATH yet. Close the terminal, open a new window and try again. If it still fails, run the PATH command the installer printed under Setup notes, or follow the fix in Anthropic's terminal guide, then check claude --version.
Academl's Terminal Agent for Absolute Beginners course has 6 lessons of 10 to 15 minutes, about 76 minutes in total, with no coding. The first 2 lessons are free without sign-up; signing up gives 7 days of full access with no card, then Pro is $9 a month or $79 a year.
Open the first 2 lessons of Terminal Agent for Absolute Beginners for free: why the phone chat is 10% of what AI can do, and what a terminal agent is. For the remaining 4 lessons with tasks and a certificate, sign up for 7 days of full access with no card, then Pro is $9 a month or $79 a year.