Claude is an AI assistant from Anthropic that you can talk to in plain language - one that helps you write text, answer questions, develop ideas, and, especially in this tutorial series, read, write, and improve real code. What makes Claude so useful isn't just that it answers questions, but that it can carry out several steps on its own, one after another: reading files, making changes, running commands, checking the result - and fixing things up if needed.

Before this series goes into depth, a quick overview is worthwhile: Claude doesn't come in just one form, but in several, depending on what you're trying to do. This chapter shows you every way to access it and helps you understand what each one is for - and why this series clearly focuses on one of them in particular: Claude Code.

Claude.ai in the browser

Claude.ai is the website where you can chat with Claude directly in your browser - no installation required, just a browser tab. You type in a question or task, Claude answers, and you can keep the conversation going: asking follow-up questions, refining results, adding new angles. Claude.ai is great for anything that can be worked out through conversation: writing or revising text, having concepts explained, brainstorming ideas, analysing documents you upload.

Claude.ai also has "Projects": a kind of folder where you bundle several conversations and uploaded files around one topic, so Claude keeps the same context in view across multiple conversations. The next chapter shows exactly how that works.

Claude Code: the CLI tool

Claude Code is Claude's home in the terminal - a command-line tool ("CLI" stands for "Command Line Interface") that you start inside a project folder, where it works directly with your actual code. The key difference from Claude.ai: Claude Code can independently read files in your project, change them, create new files, and run commands like tests or build scripts - always with your permission, as you'll learn in detail in block 2 of this series.

That makes Claude Code the right tool for anything that happens directly in the code: building a new feature, fixing a bug, understanding an existing project, writing tests, creating git commits. That's what most of this tutorial series is about - starting in block 3 you'll even build a small example project with it, step by step.

Desktop and mobile app

Claude is also available as a standalone desktop app (for Windows and Mac) and as a mobile app (for iOS and Android). Both offer essentially the same chat experience as Claude.ai in the browser, just as an installed app instead of a website - handy if you want Claude within reach without opening an extra browser tab, or if you want to ask a quick question on your phone while out and about.

The API: building Claude into your own software

The fourth way to access Claude is aimed at developers who want to build Claude into their own programs and websites: the API ("Application Programming Interface"). With it, a program sends requests to Claude and gets the answers back automatically - for example for a chatbot on your own website, or a feature inside an app. The API isn't a topic of this series; it's mentioned here purely for completeness, so you know the difference if you come across the term elsewhere.

All four access points compared

• Claude.ai (browser): conversations, writing, research, uploading and discussing files - no installation.
• Claude Code (CLI): working directly inside a project folder, reading/writing/running code - the focus of this series.
• Desktop/mobile app: the same chat as Claude.ai, as an installed app instead of in the browser.
• API: building Claude into your own applications programmatically - for developers building their own products.

For the rest of this series, the focus stays clear: chapter 2 takes a closer look at Claude.ai in the browser, and after that almost everything revolves around Claude Code - installation, tools, commands, and how to put it to meaningful use in a real project.