This tutorial shows you how to use Claude Code from zero: what it is, which plan you need, how to install it, and the workflow that actually gets features shipped. If you're getting started with Claude Code today, this is the shortest path from install to a productive first session.
What is Claude Code?
Claude Code is Anthropic's AI coding agent. Unlike a chat window where you copy and paste snippets, Claude Code works directly on your machine: it reads your files, edits them in place, runs commands and tests, and commits when something works. You describe the goal; it plans the steps and carries them out.
You can use Claude Code in several places: the terminal (CLI), the Code tab of the Claude desktop app, the VS Code extension, and on the web. They share the same engine, so the habits in this guide carry over between them.
How much does Claude Code cost?
Claude Code is included with paid Claude plans. The free plan doesn't include it. Individual plans as of October 2026:
| Plan | Price (USD) | Claude Code |
|---|---|---|
| Free | $0 | Not included |
| Pro | $20/month, or $17/month billed annually | Included |
| Max | From $100/month | Included, with higher usage limits |
Pro is enough to learn and to work through this tutorial. If you start hitting usage limits every day, move up to Max. Team and Enterprise plans include Claude Code too. Check claude.com/pricing for current prices and limits.
How to install Claude Code
Pick one of the two options. If you've never used a terminal, start with the desktop app.
Option 1: The Claude desktop app (no terminal needed)
- Download Claude for macOS or Windows (a Linux beta is also available) and sign in with your paid account.
- Click the Code tab at the top of the window.
- Leave the session on Local, click the folder button, choose Open folder…, and pick the project you want Claude to work in.
- Type what you want in the prompt box and press Enter.
Option 2: The terminal
Run the official installer for your system.
macOS, Linux, or WSL:
curl -fsSL https://claude.ai/install.sh | bash
Windows PowerShell:
irm https://claude.ai/install.ps1 | iex
Homebrew users can run brew install --cask claude-code instead. The native installer updates itself in the background; Homebrew installs need a manual brew upgrade. (The older npm install -g @anthropic-ai/claude-code route still works if you have Node.js 22 or later, but the native installer is the recommended path.)
Open a new terminal window and confirm it worked:
claude --version
Then go to your project folder and start a session:
cd my-project
claude
The first run opens your browser to log in. After that, your login persists.
Your first Claude Code session, step by step
Whichever option you chose, a good first session looks like this:
- Ask Claude to explain the project. "Give me an overview of this codebase: what it does, how it's structured, and how to run it." You learn the repo and Claude loads the context it needs.
- Ask for one small, concrete change. "Add a loading spinner to the submit button in
src/components/Form.tsx." Small requests are easy to review. - Review what it did. Read the diff before accepting it. Ask why it made a choice if something looks off.
- Have it run the thing. "Run the tests" or "start the dev server and check the page loads." Claude sees the output and fixes what breaks.
The mental model that matters
Treat Claude Code as a junior engineer with infinite patience and your full repo access. That framing changes how you prompt:
- Don't dictate every step. Describe the goal and the constraints.
- Reference real files. Paste paths, function names, line numbers — anything that anchors the model in your codebase.
- Push back. If the first attempt is wrong, say what's wrong specifically. Don't restart from scratch.
A Claude Code workflow that actually ships
Here's the loop most experienced users settle into:
Step 1: Plan before code
Before touching files, ask Claude to outline the change. "Read these three files and tell me where the auth check lives" beats "fix the auth bug" by a wide margin. Plan mode is built for this: Claude researches and proposes a plan, and doesn't edit anything until you approve it.
Step 2: Make small, reviewable diffs
Ask for one slice at a time. A 50-line diff you actually read is worth more than a 500-line diff you skim.
Step 3: Run the thing
After every change, run the test suite or the dev server. Claude Code can do this for you — and it will see the output, which is the whole point.
Level up: Skills, MCP, and subagents
Once the basic loop feels natural, three features compound your productivity:
- Skills are reusable instruction packs written in markdown. Package a workflow you repeat — scaffolding a project, writing release notes — and Claude follows it whenever the task comes up.
- MCP servers connect Claude Code to outside services such as GitHub, Vercel, Notion, or your database, so it can act on them directly instead of asking you to click through a dashboard.
- Subagents handle side tasks in their own context. For independent searches ("find every place we call
getUser") or to protect your main conversation from a 2000-line file dump, spawn a subagent. The Explore agent in particular is great for read-only investigation.
Common pitfalls
- Vague prompts. "Make this better" produces nothing useful. "Reduce the bundle size of
src/app/page.tsx, target under 150kb gzipped" produces a real diff. - Skipping the verification step. A green test suite is the only thing that confirms a fix works. Trust the suite, not the model's confidence.
- Letting the context bloat. Long sessions accumulate noise. Start fresh sessions for unrelated tasks.
FAQ
Is Claude Code free?
No. Claude Code requires a paid plan: Pro or Max for individuals, Team or Enterprise for organizations, or a Claude Console account billed by API usage. The cheapest option is Pro at $20 a month.
Can I use Claude Code on Windows?
Yes. Install the Claude desktop app for Windows and use the Code tab, or run the PowerShell installer above. Claude Code runs natively on Windows 10 (version 1809) or later, and inside WSL.
Do I need to know how to code?
No, but it helps. You can build and ship a small project by describing what you want and approving Claude's changes. Knowing how to read code lets you review its work and give more precise instructions.
What's the difference between Claude Code and regular Claude chat?
Chat can only tell you what to do and hand you code to copy. Claude Code does the work: it edits the files on your machine, runs commands, reads the results, and iterates until the task is done.
Next steps
The official Claude Code docs cover slash commands, hooks, Skills, and MCP integrations in detail — that's where the real productivity gains compound. Bookmark them and come back as you grow into the tool.