How to learn Claude Code for free, in the right order
Anthropic ships about twenty free courses. Most people pick the wrong one first. Here is the order I would take them in, depending on what you want.
People keep asking me where to start with Claude Code. For a long time I did not have a short answer.
Now I do. Anthropic has quietly built a full curriculum. About twenty free courses, with tests and certificates, and it is no longer a pile of intro videos. There are separate tracks for regular users, for developers, for people who want Claude Code specifically, for the API, for MCP, for cloud integrations.
The catch: you should not take them in the order they are listed.
The main idea is simple. Pick the track that matches your goal, skip the rest, come back later if you need to.
If your goal is Claude Code itself
This is the order I would take:
1. Claude Code 101 How a coding agent actually works: exploring a project, planning, writing code, committing changes. Do not skip this even if you have used Claude Code for a while. It fixes the mental model.
2. Claude Code in Action The daily workflow. Files, commands, context, GitHub, a full development loop from task to pull request.
3. Introduction to Agent Skills Reusable instructions for the things you do every week: testing, code review, deployment. This is where Claude Code stops being a chat and starts being a system.
4. Introduction to Subagents How to hand off parts of a task to isolated agents so your main context does not drown.
5. Introduction to MCP Only after the first four. Connecting external tools, services and data makes sense once you know what the agent does with them.
If you want to build your own AI products
Different track, different order:
Claude Platform 101 → Building with the Claude API → Introduction to MCP → Model Context Protocol: Advanced Topics
The API course is the one serious commitment in the whole catalog. It runs over eight hours. Everything else is twenty minutes to an hour.
If you are new to AI altogether
Start even simpler:
Claude 101 → AI Capabilities and Limitations → AI Fluency
What to skip
The Amazon Bedrock and Google Vertex AI courses are only worth it if you actually deploy on those platforms. If you do not, they will teach you nothing you can use this month.
Why I like these over the usual tutorials
They are official. That matters more than it sounds.
Most Claude Code tutorials online are a list of commands and tricks. These courses teach the working model the product team itself has in mind: how to set a task, how to explore a codebase, how to plan changes, how to manage context, when to reach for skills, subagents and external tools.
You learn the tool the way its authors think about it. Tricks you can pick up later. The model of work is the part that is hard to get from YouTube.
All of it is free. There are quizzes and certificates if you care about those.
If you want to learn Claude Code and your English is decent, this is one of the best entry points available right now. Take the five courses in the first list, in that order, and do one real task of your own between each of them.
Full catalog: claude.com/resources/courses