Setting Up Claude

Claude is made by a company called Anthropic. People tend to reach for it when they want careful writing or a problem worked through slowly. This page walks you through setting it up.

If you haven't decided whether Claude is your tool, the Pick a Tool page compares it to ChatGPT and Gemini without crowning a winner. The three are close, and everything you learn here works the same in all of them. If you're here, you've either chosen Claude or you're trying it alongside another tool. Either is a fine reason to keep reading.

Take it slowly. The whole setup is about fifteen minutes if you stop to breathe. You can always close the page and come back tomorrow.

Getting to Claude

Open a new tab in your browser: Chrome, Safari, Edge, Firefox, whichever one you usually use.

Click in the long white bar at the top (the address bar, where website addresses show up). Type this exactly:

claude.ai

Press Enter on your keyboard.

What you should see: a clean page with the word "Claude" near the top, and a big button that says either Sign up, Get started, or Try Claude. There's usually a small drawing or pattern in the background; that's normal.

If you see something different (a broken page, a search engine, a different website), you probably typed the address wrong. Go back to the address bar and try again. It's claude.ai (not .com, not .org). Nothing fancy.

Making your account

Click the Sign up (or Get started) button.

Claude will offer you two ways to sign up:

• With your Google account. If you already have Gmail, this is the easiest. Click the Google option, pick your Gmail, approve a few things, done. No new password.
• With an email and new password. If you don't want to link it to Google, type in your email, pick a password, and Claude will send you a verification email.

Pick whichever feels easier. The Google way takes about ten seconds. The email way takes about two minutes because you'll need to switch to your email, open the verification message, and click the link inside it.

What you should see after signing in: a mostly-empty screen with a big text box in the middle that says something like "How can I help you today?" or "Message Claude..." That's the main Claude chat. You're in.

About the plans

You're now on the Free plan automatically. You didn't pay anything. Nothing is going to charge you.

Here's the honest version of what each plan does:

• Free: enough chats per day to get a feel for it. You'll hit a limit after an hour or two of serious use. Projects are not available on Free.
• Pro: around $20 USD a month, roughly the same as the paid tier for ChatGPT and Gemini. This turns on Projects (a saved workspace, covered later in this guide), longer conversations, and a good deal more usage. This is what you'll want eventually.
• Max: a higher tier for heavy daily users. Most people don't need it. Skip it.
• Team: for groups sharing one account. Skip it unless that's you.

Turning on the good features

Claude has a few features that make it much more capable. They're turned off by default on some accounts. Turn them on now so they're ready when you need them.

Find the settings. Look at the very bottom-left corner of the page. You'll see either your name, your initials, or a little profile circle. Click it. A small menu pops up above it.

From that menu, click Settings.

What you should see: a settings screen with a list of categories on the left: things like "Profile," "Appearance," "Feature preview," "Privacy," and so on.

Click Feature preview (sometimes called Beta features). This is a list of features Claude is trying out. Each one has an on/off switch next to it.

Turn these on by clicking the switch next to each one:

• Artifacts: lets Claude put a document, table, or little webpage in a side panel so you can look at it next to the chat. Essential.
• Analysis tool: lets Claude calculate things with numbers and read spreadsheets. Don't use it yet, but have it ready.
• Create files or File creation: lets Claude make PDFs, spreadsheets, and slides you can download. Useful.

If you don't see one of these: don't panic. Anthropic renames and moves features every few months. Just turn on whatever is available. If they're all on by default when you look, even better.

When you're done, click out of Settings (there's usually an X in the corner, or just click outside the settings box).

A quick tour of the Claude screen

Before you leave, get familiar with what you're looking at. There are really only three parts to notice. Here's roughly what the page looks like, with the important spots labeled.

The big text box in the middle. This is where you type your questions. Below the box are small buttons for attaching files, taking a picture, or recording your voice. You don't need any of those today.

The left sidebar. Down this column you'll see:

• A New chat button at the top.
• A list of your past conversations, empty for now, will fill up as you use it.
• A section called Projects. This is the feature we'll come back to later. Don't click it yet.

The model dropdown. Near the top of the chat, there's a small dropdown that probably says Claude Sonnet or Claude Opus followed by a number. Those are different versions of Claude. Sonnet is faster. Opus is slower but thinks harder. You don't need to change this; whatever's selected by default is fine.

One small test

Before you close the page, type this into the chat and press Enter:

Hi Claude. I'm just learning how to use AI properly. Can you explain, in plain English for someone who isn't technical, what makes you different from ChatGPT? Keep it to two short paragraphs.

Read what comes back. Notice the tone: Claude tends to be a bit more careful and explain-y than ChatGPT. That's useful when the answer matters. Less useful when you just want a one-line response (we'll fix that on the Prompting page).

A word on what Claude remembers

Two different things are easy to mix up. Inside a single conversation, Claude holds everything you've said so far (its context window, explained on the Basics page), and forgets it all when you close the chat. Separately, Claude can carry some memory across chats, so it recalls a few things about you from one conversation to the next; look for a memory or personalization setting, and note that you can view and clear what it has kept. The exact wording moves around, so search Settings for "memory" if you don't see it.

For the work in this guide, the reliable way to make Claude remember what matters isn't memory at all. It's a Project, where your files and instructions live permanently and get loaded into every chat automatically. That's the next thing you'll build.

Where to check this

Claude's plans, features, and settings change over time. For current details, Anthropic's own help centre at support.anthropic.com is the reliable source. Search there for a feature name if a setting has moved since this was written.