How to Use Claude: A Step-by-Step Guide for Beginners (2026)

To use Claude, go to claude.ai or download the Claude app, sign up for free with your email address, and type what you need into the message box in plain English, the way you’d ask a coworker. Then keep the conversation going: ask follow-up questions, add files with the + button, and switch on web search when you need current information.

This guide walks through everything a beginner needs, in order: signing up, writing your first prompts, uploading files, picking a model, web search and Research, Projects, memory, voice mode, finished files and artifacts, connecting your apps, and what the Free, Pro and Max plans include. Every feature and limit below was checked against Anthropic’s own Help Center and pricing page. Writing code? Our separate guide covers how to use Claude Code.

Claude at a glance

Checked against the Claude Help Center and the claude.com pricing page on October 8, 2026.

Step 1: Sign up and pick where you’ll use Claude

Go to claude.ai in any browser and sign up with your email address. The Free plan doesn’t need a payment card. You must be at least 18 and in one of Anthropic’s supported locations.

You can use the same account in three places:

• On the web: claude.ai in any browser. Nothing to install.
• On your computer: the Claude desktop app for macOS 11 or later, Windows 10 or later, and Linux (in beta), from Anthropic’s downloads page. It’s the app you need for features that work with files and folders on your computer.
• On your phone: search for “Claude by Anthropic” in the App Store (iOS and iPadOS 18 or later) or on Google Play (Android 8.0 or later). The phone app is where voice mode works best.

Step 2: Write your first prompt

A prompt is just your message. Anthropic’s own advice is to talk to Claude naturally, the way you’d talk to a coworker or a friend, and to be specific. A prompt can be a one-line question or a multi-step project.

Three habits make the biggest difference for beginners:

• Give context. Who is it for, why do you need it, and what does good look like? “Write a thank-you note” gets a generic note. “Write a short thank-you note to my neighbor who watered my plants for two weeks, warm but not gushing” gets one you can send.
• Ask for a format. Bullets, a table, three options, under 100 words: say exactly how you want the answer laid out.
• Keep going. The first answer is a draft. Reply with “shorter,” “more casual” or “give me two more options” instead of starting over.

Not sure what to ask? Paste one of these and swap in your details:

• “Explain how a 401(k) match works like I’m new to this, then ask me 3 questions to check I understood.”
• “Here’s an email I need to answer. Draft a polite reply that says no to the meeting but offers a 15-minute call instead.”
• “Plan five weeknight dinners for two using chicken, rice and eggs, under 30 minutes each, with one shopping list.”
• “I’m stuck on [problem]. Ask me five questions about it before you suggest anything.”

For 20 more ready-made prompts, grab our copy and paste AI prompts, and for the reasons they work, read the complete guide to AI prompting.

Step 3: Upload files and photos

Claude is at its most reliable when it works from your material instead of its memory. To add a file, click the + button in the lower left of the message box, choose Add files or photos and pick the file. You can also drag files into the chat or paste an image from your clipboard.

Two tips from Anthropic’s help center: when you ask about a specific page, use the page number your PDF viewer shows rather than the one printed on the page, and split very large documents into smaller sections. Photos work too: a screenshot of an error message, a confusing bill or a chart buried in a report are all fair game.

Step 4: Choose a model and how hard Claude thinks

The model menu next to the send button controls three settings: which Claude model you’re using, how much effort it puts into each answer, and whether it thinks before responding. You can change any of them in the middle of a conversation.

• Model. Claude comes in four families: Haiku is the fastest, Sonnet is a strong all-rounder, Opus is more capable, and Fable is Anthropic’s newest top tier for long-running work. The Free plan includes Sonnet and Haiku, and Pro adds Opus. Fable models are on paid plans: on Pro they run on pay-as-you-go usage credits, and on Max you can spend up to half your weekly limit on them.
• Effort. Low and Medium suit routine tasks and stretch your usage further, High is the best overall balance, and Extra high and Max are for the hardest work. Higher effort uses your limits faster.
• Thinking. Claude works through the problem before answering and shows its reasoning in an expandable Thinking section. On Anthropic’s newest models, thinking is always on.

For everyday questions and writing, the defaults are fine. Raise the effort for complex analysis, planning, math or multi-step problems.

Step 5: Turn on web search (and Research for big questions)

Every Claude model learns from data up to a cutoff date; Anthropic’s newest models were trained on data up to June 2026. For anything that changes, like prices, news, product details or opening hours, have Claude search the web: click + in the message box and turn on Web search, or start your prompt with “Search the web for…”. Answers that use search include citations you can click to check the source.

If your account has the newer, combined Claude experience (rolling out first to Pro and Max), there’s no web search toggle: Claude searches the web on its own when it helps.

For bigger questions, Research (paid plans) runs many searches that build on each other and returns a cited answer in minutes. Turn it on from the + menu, or type /deep-research in the newer experience. It uses your limits faster than a normal chat. Whatever you use, click through on anything important: Anthropic’s own help center says Claude can occasionally produce incorrect or misleading responses. Our guide to improving AI accuracy has seven fixes that help.

Step 6: Use Projects for work you come back to

A project is a workspace with its own chats, files and instructions. Upload your reference documents once, add instructions like “use a formal tone” or “answer as a career coach,” and every chat inside the project starts with that context. Projects are on every plan, and the Free plan allows up to five.

1. Open claude.ai/projects (or click Projects in the sidebar) and click + New Project.
2. Give it a name and description. Claude can’t see these, so put anything it needs in the project itself.
3. Click + in the project knowledge area and upload your documents, text or code.
4. Click Set project instructions, describe how Claude should behave, and save.
5. Start a chat inside the project.

One gotcha: chats in the same project don’t see each other unless the information is in the project knowledge. On paid plans, big projects switch automatically to a retrieval mode that expands capacity by up to 10 times. Projects also save usage, because project files are cached and count less against your limits when you reuse them.

Good first projects: a job search (your résumé plus the roles you want), a side business, a class you’re taking, or a report you write every month. Our guide to the parts of Claude most people never find shows how Projects fit with Claude’s other modes.

Step 7: Let Claude remember you (or not)

Claude has three ways to carry context from one chat to the next:

• Memory: on by default on Free, Pro and Max. Claude saves useful details as you chat, like your role, your projects and how you like things written, and uses them later. Say “remember this” to save something on purpose. See and edit everything it remembers in Settings > Memory. Each project keeps its own separate memory.
• Instructions for Claude: account-wide instructions in Settings that apply to every conversation, such as “I’m a nurse. Keep answers practical and short.”
• Chat search (paid plans): ask “What did we decide about the kitchen remodel?” and Claude searches your past chats.

Want a clean slate? Click the ghost icon in the upper right of a new chat to start an incognito chat. It isn’t saved to your chat history or memory, and it isn’t used to train Anthropic’s models. To keep a chat but leave memory out of it, turn off Memory in the + menu before you send your first message. By default Claude doesn’t save sensitive topics such as health or religion to memory, and it never saves things like government ID or financial account numbers.

Step 8: Talk to Claude with voice mode

Voice mode lets you speak to Claude and hear it answer. It’s in beta on every plan, on the web, the desktop app and mobile, and it’s built to work best on your phone. Tap the sound wave icon in the message box, choose a voice and start talking. Hands-free mode replies when you pause; switch to push-to-talk in noisy places. Voice conversations count toward your usage limits.

Talking often gets you better answers because you naturally give more context. Here’s why talking to AI beats typing.

Step 9: Get finished files, docs and decks

Claude doesn’t only answer in text. On every plan it can run code to create real files you can download: Excel spreadsheets with working formulas, PowerPoint presentations, Word documents and PDFs. It can also make artifacts: designs, decks, documents, dashboards and small interactive tools that open beside the conversation, where you can edit them, come back to them and share them.

Templates for Docs, Slides and Design (all in beta) are on every plan. Decks export to PowerPoint, PDF or Google Slides, and docs export to Word, PDF, Markdown or Google Docs. Artifacts need code execution and file creation turned on in Settings > Capabilities, which is the default on Free, Pro and Max.

What Claude won’t do is generate photos or illustrations; it builds charts, diagrams and interactive visuals instead. For AI images, use ChatGPT or Gemini (here’s how the three compare).

Try: “Turn these meeting notes into a one-page Word document with a table of action items, owners and due dates.”

Step 10: Connect your apps

Connectors let Claude read from, and act in, services you already use, like Google Drive, Gmail, Slack or Microsoft 365. Go to Customize > Connectors, pick a service and sign in. Once it’s connected, Claude can bring it into a conversation when it fits your request, and you can switch connectors on or off for a chat from the + menu. Connectors work on every plan, but the Free plan lets you connect only one.

Only connect services you trust: you’re giving Claude permission to access, and possibly change, data based on your own account permissions.

Hand off bigger tasks (paid plans)

On paid plans, Claude can take on whole tasks, not just answer questions. Anthropic is merging regular chat and Claude Cowork into one Claude experience, rolling out first to Pro and Max: you describe the outcome, and Claude works through the steps itself, searching the web, reading your files, running code and handing back finished documents, spreadsheets or presentations. Longer tasks keep running in the cloud if you close your laptop, and you can schedule recurring work, like a summary of your inbox every Monday morning.

A permission setting in the message box decides how independently Claude works: Manual (the default) asks before each action, and Auto keeps going with automated safety checks. In Claude Desktop you can give Claude a folder on your computer, and Claude can browse websites with the browser built into the desktop app or with the Claude in Chrome extension (paid plans only). Stay close for anything with real consequences, like sending messages or changing important files.

If you write software, Claude Code is the tool built for that, and it’s included with every paid plan. Start with our beginner’s guide to Claude Code.

Free vs Pro vs Max: what each plan includes

Start free. Upgrade when you keep hitting the five-hour limit on work you do every day, or when you need Research, Opus or Claude Code. Pro can be billed monthly or yearly, while Max is monthly only. Anthropic says prices vary by region and that it doesn’t offer standing discounts, only occasional limited-time promotions. A full Claude Free vs Pro comparison is coming to this site soon.

How Claude’s usage limits work

Claude doesn’t give you a fixed number of messages. Every plan has a session limit that resets every five hours, and paid plans add a weekly limit on top. How fast you use it depends on how long your messages and files are, how long the conversation has run, the model and effort level, and tools like Research and web search. Claude on the web, desktop and mobile and Claude Code all draw from the same pool, and Claude tells you when you’ve reached a limit.

To stretch your usage:

• Start a new chat when you change topics, because long conversations use more with every message.
• Put documents you reuse into a project, where they’re cached.
• Batch related questions into one message.
• Use a lower effort level for routine tasks, and turn off connectors and web search you don’t need.
• Check where you stand in Settings > Usage, which shows your session and weekly limits on paid plans.

When you hit a limit you can wait for the reset, upgrade, or, on paid plans, turn on usage credits to keep working at standard API rates.

Stay safe and keep your data private

• Check what matters. Claude can be fluent, specific and wrong. Use web search for anything current, click the citations, and verify anything high-stakes. These 3 prompts catch most hallucinations.
• Choose your training setting. On Free, Pro and Max, Anthropic uses your chats to improve its models only if you allow it, or if a chat is flagged for a safety review. Change it in Settings > Privacy under Help improve our AI models. Incognito chats are never used for training.
• Leave out secrets. Don’t paste passwords, account numbers or other people’s private details, and follow your employer’s AI policy for work material. Here’s how to check what AI already knows about you.

10 things to try with Claude this week

1. Upload a long PDF, like a lease or a policy, and ask for five bullets plus anything you need to do.
2. Paste an email you’ve been avoiding and ask for a reply in your voice. More in how to write better emails with AI.
3. Turn messy meeting notes into decisions, owners and due dates (here’s the full routine).
4. Ask it to explain a confusing bill or form at your level, then quiz you on it.
5. Create a project for your job search with your résumé inside.
6. Use voice mode on a walk to think through a decision out loud.
7. Upload a spreadsheet and ask which categories changed the most (see make spreadsheets talk to you).
8. Ask for a 10-slide deck from your notes and export it to PowerPoint.
9. Have it build a study plan for something you’re learning, or use our free AI study plan generator.
10. Ask it to stress-test a plan: “Be brutally honest about what I’m underestimating.”

Want more ideas? See what AI can do today for 12 everyday uses that work in any assistant. And if you build with Claude at work, Anthropic now offers certification exams; our Claude certification guide explains who can take them and how to prepare.

Related guides

• How to Use Claude Code: A Step-by-Step Beginner’s Guide (2026)
• The AI Beginner’s Playbook: 20 Copy and Paste AI Prompts That Actually Work
• The Complete Guide to AI Prompting (for People Who Aren’t Engineers)
• How to Improve AI Accuracy: 7 Fixes That Actually Work
• You’re Only Using 30% of Claude AI — The Part of Claude Most People Never Find