How to Use Claude: A Practical Beginner's Guide (October 2026)

To use Claude, open claude.ai in a browser (or the desktop or mobile app), sign in, type what you need into the chat box in plain language, and press send. The real skill is in what comes next: giving Claude enough context, attaching the right files, and using Projects, Artifacts and memory so you don’t repeat yourself every time.

This guide walks through all of it, based on Anthropic’s own Help Center and prompting documentation. Features and plan details change often, so treat everything here as accurate as of October 2026.

Where you can use Claude

Anthropic lists several official ways to reach Claude in its interfaces article:

According to the Get started article, you need to be at least 18 and in a supported location. Stick to these official sources; there is no need for any third-party download.

Your first conversation, step by step

1. Go to claude.ai or open the app and sign in or create an account.
2. Type your request into the chat box. Anthropic recommends talking to Claude “naturally and conversationally,” the way you would with a coworker.
3. Click the submit button.
4. Read the answer, then reply with follow-ups to refine it. You don’t need to start over when something is off; just say what to change.
5. Click the ”+” button in the lower left, or type ”/”, to see extra options and commands.

Checking which model you’re using

The current model appears below the chat box on web and desktop, or at the top of the screen on mobile. Click it to switch models. From the same picker you can adjust how much effort Claude puts into a response and whether it uses extended thinking. Anthropic’s newest flagship, Claude Opus 5.5, was released on September 22, 2026. Which models you see depends on your plan; the pricing page lists “more Claude models” as a Pro feature.

A sensible default: use the standard model for everyday questions, and switch to a more capable model or turn up effort for hard reasoning, long documents or coding.

How to write prompts that work

Anthropic’s prompting best practices are written for developers, but the core advice applies just as well in the chat app.

Be clear and specific

Anthropic suggests thinking of Claude as “a brilliant but new employee who lacks context on your norms and workflows.” Its golden rule: if a colleague with no background would be confused by your prompt, Claude will be too.

Compare these two requests:

• Weak: “Write an email about the meeting.”
• Better: “Write a short, friendly email to my team moving Thursday’s 2pm planning meeting to Friday at 10am. Ask them to reply if Friday doesn’t work. Keep it under 100 words.”

Say what format you want (bullet list, table, a single paragraph), how long it should be, and who it’s for. If you want Claude to go “above and beyond,” ask for it explicitly rather than hoping it guesses.

Explain why

Adding the reason behind an instruction helps Claude generalize. Anthropic’s example: instead of “never use ellipses,” explain that the text will be read aloud by a text-to-speech engine that can’t pronounce them.

Show examples

Examples are one of the most reliable ways to steer tone and structure. If you have a past email, product description or report you like, paste it in and say “match this style.” Anthropic recommends three to five varied examples for best results.

Give Claude a role and structure long prompts

A single sentence like “You are an experienced copy editor” focuses the response. For long prompts that mix instructions, background and source material, put each part in its own clearly labeled section so Claude doesn’t confuse your instructions with the text you want it to work on.

Iterate

The Help Center’s own tips are simple: start with easy questions, be specific, refine with follow-ups, and try different kinds of tasks. Treat the first answer as a draft.

Uploading files and images

Claude can read documents and images you share. Per the upload help article:

1. Click ”+” in the lower left of the chat box.
2. Choose “Add files or photos.”
3. Pick your files, or drag and drop them into the chat. You can also paste images from your clipboard.

Supported documents include PDF, DOCX, CSV, TXT, HTML, ODT, RTF, EPUB, JSON and XLSX (XLSX requires the code execution and file creation setting). Images can be JPEG, PNG, GIF or WebP. In a chat, the limit is 20 files of up to 500MB each, and PDFs up to 1,000 pages. Claude only analyzes charts and visuals in PDFs of 100 pages or fewer; longer PDFs are processed as text. Some older copies of this article listed lower limits, so check the live page if a file is rejected.

Good uses: “Summarize this contract and list anything unusual,” “Turn this spreadsheet into three key takeaways,” or “What’s wrong with the wiring in this photo?”

Use Projects for recurring work

If you keep pasting the same background into new chats, use a Project. A Project is a workspace with its own files and instructions that apply to every chat inside it. The Projects article says they’re available to all users, with free accounts capped at five projects. (The pricing page lists Projects under Pro, so the exact free limits may shift.)

1. Open Projects in the left sidebar, or go to claude.ai/projects.
2. Click ”+ New Project” and give it a name and description. Claude can’t see these; they’re just for you.
3. Click ”+” in the knowledge area on the right and upload reference documents, text or code. Project files are limited to 30MB each.
4. Click “Set project instructions,” describe how Claude should respond (tone, audience, format), and save.

One catch: chats inside a Project don’t share context with each other unless the information is in the project knowledge or instructions. Each project also keeps its own memory, separate from your other projects.

Let Claude build things with Artifacts

When you ask for something substantial and reusable, such as a document, slide deck, dashboard, diagram or small interactive tool, Claude opens it as an Artifact in a panel beside the chat. According to the Artifacts article:

• Artifacts need the “Cloud code execution and file creation” capability, which Free, Pro and Max users turn on under Settings > Capabilities.
• Basic Artifacts and the Docs, Slides and Design templates are available on all plans. Dashboards (beta) and connecting Artifacts to your apps need Pro or higher.
• Ask Claude for changes in chat, or highlight text in a doc and click “Edit with Claude.”
• Docs export to Word, PDF, Markdown or Google Docs; decks export to PowerPoint, PDF or Google Slides.
• Artifacts are private until you share them, and sharing isn’t available in the iOS and Android apps.

Try prompts like “Make a one-page project brief as a doc” or “Build a simple budget tracker I can use.”

Manage memory and privacy

Claude’s memory saves topics such as your role, projects and preferences and uses them in later chats. It’s on by default for Free, Pro and Max users; on Team and Enterprise, an owner has to enable it. Sensitive topics like health or politics aren’t saved by default.

• Review or edit: Settings > Memory lists everything under “Topics.” You can edit or delete any of them, or just tell Claude in chat what to remember or forget.
• Turn it off: switch off “Generate memory from chats” in Settings > Memory. You can pause (keeping existing memories) or reset (deleting them permanently).
• One chat only: click ”+” before your first message and turn off Memory.
• Incognito: click the ghost icon in a new chat. These chats aren’t saved to history or memory, and they’re available on every plan.

Paid plans also get chat search, which lets Claude look up past conversations (“What did we decide about the logo?”). You can switch it off with “Search and reference chats” in Settings > Memory.

Know the limits

Free plans have a session limit that resets every five hours, and Anthropic says message allowances vary with demand. Claude also tells you when a prompt is too large for its context window. If you run into limits often, paid plans start with Pro at $20 a month billed monthly ($17 a month billed annually) as of October 2026, with Max offering 5x or 20x Pro’s usage.

Finally, Claude can still be wrong. Its training data has a cutoff, so for current events, prices or anything important, check facts against the original source. Web search, which is included even on the Free plan, helps, but doesn’t replace your own judgment.