Anthropic Academy Guide: All 13 Free Claude Courses

Anthropic launched Anthropic Academy in March 2026. Thirteen courses. All free. Self-paced. Official certificates on completion. Everything that was scattered across blog posts, docs, and GitHub repos -- consolidated into one learning platform.

This article covers every course, its prerequisites, and recommended paths based on your role. When courses get added or updated, this page updates too.

All 13 Courses at a Glance

Two groups. Six courses require no coding -- they work for everyone. Seven are developer-focused, requiring Python or CLI experience.

For Everyone: Getting Started with Claude

Claude 101

If you take only one course, take this one. It teaches you to use Claude for everyday work: writing emails, organizing data, analyzing documents, drafting content. No technical background needed. Just how to talk to an AI and get real work done.

This is a starting point. After completing it, you will know what Claude can do and which direction to go deeper.

The AI Fluency Series

A course series co-developed by Anthropic and academic partners. Built around the "4E Framework": Effective, Efficient, Ethical, Safe.

Framework & Foundations is the universal version. It teaches thinking frameworks, not button clicks. Where AI's capability boundaries lie. Which tasks to delegate to AI. When human judgment cannot be replaced. You walk away knowing not just how to use Claude, but when not to. Taught by Prof. Joseph Feller (University College Cork) and Prof. Rick Dakan (Ringling College). Certificate on completion.

The other four are audience-specific:

• For Students: Academic research, career planning, learning strategies
• For Educators: Integrating AI into instruction, curriculum design, assessment strategies
• Teaching AI Fluency: How to teach others to use AI. Scenario design, learning outcome evaluation
• For Nonprofits: Co-developed with GivingTuesday. Tailored for resource-constrained organizations

For Developers: From API Calls to Agent Architecture

This is where the technical depth lives. Seven courses covering the full Claude development stack, from your first API call to deploying production-grade agent systems.

Building with the Claude API

Prerequisites: Python, JSON basics

The required course for developers. Starts from your first API call. Builds up to agent system architecture:

• API setup and authentication (key management, request format)
• Multi-turn conversations and system prompt design
• Prompt engineering (XML tag structuring, few-shot learning)
• Tool use: letting Claude call external functions you define
• RAG (Retrieval-Augmented Generation): giving Claude access to your data
• Extended thinking mode and prompt caching
• Agent architecture: parallelization, routing, multi-step orchestration

Backend or full-stack engineer building Claude into a product? Start here.

Claude Code in Action

Prerequisites: Terminal experience, Git basics

Claude Code is Anthropic's agentic coding tool. It reads your codebase, edits files, runs commands. This course teaches daily workflow integration:

• How the tool system interacts with your codebase
• Context management strategies (helping it understand your project)
• Custom commands (automating repetitive tasks)
• MCP server integration (connecting external tools like browser automation)
• GitHub workflows (automated code review)
• When to use thinking mode vs. planning mode

After this course, follow up with our hands-on First Hour with Claude Code tutorial to practice on your own project.

Introduction to Agent Skills

Prerequisites: Basic Claude Code experience

Agent Skills extend Claude Code through Markdown files (SKILL.md) that it automatically applies at the right moments. The course covers:

• How Skills differ from CLAUDE.md, hooks, and subagents
• Writing a SKILL.md from scratch (frontmatter format, trigger descriptions)
• Directory structure design (managing context window consumption)
• Advanced configuration: restricting tool access, background scripts
• Team sharing (committing to repos) and enterprise deployment (managed settings)
• Wiring Skills into custom subagents

Pairs well with our Write Your First SKILL.md tutorial.

The MCP Series (2 Courses)

MCP (Model Context Protocol) is an open standard from Anthropic. It lets AI agents connect to external tools and data sources through a unified interface.

Introduction to Model Context Protocol teaches you to build MCP servers and clients from scratch in Python. Three core primitives:

• Tools: Model-controlled actions -- querying databases, making API requests
• Resources: Application-controlled read-only data -- config files
• Prompts: Pre-defined instruction templates -- "review this code"

You also learn to test and debug using the MCP Server Inspector in your browser.

MCP: Advanced Topics goes deeper into production concerns:

• Sampling: letting servers request language model calls through clients
• Transport: stdio (local) vs. Streamable HTTP (remote)
• Roots-based file access permission design
• Horizontal scaling and load balancer deployment

Take the intro first. Together, both give you everything needed to build a deployable MCP server from scratch.

Cloud Platform Integration (2 Courses)

If your team deploys Claude on AWS or GCP:

• Claude with Amazon Bedrock: Deploying on AWS Bedrock. Tool use, RAG, agent architecture
• Claude with Google Vertex AI: Same on GCP Vertex AI, plus vision, PDF processing, citations

Both overlap significantly with Building with the Claude API. Main differences: deployment platform and authentication. Already took the API course? Pick whichever cloud you use.

Learning Resources Beyond Anthropic Academy

Official Documentation

• Claude API Docs: Complete API reference -- all endpoints, parameters, error codes
• Claude Code Docs: Full Claude Code documentation, installation to advanced configuration

GitHub Open-Source Courses

anthropics/courses provides 5 sets of Jupyter Notebooks:

1. Anthropic API Fundamentals: Core API operations
2. Prompt Engineering Interactive Tutorial: 9 chapters, interactive
3. Real World Prompting: Production prompt design
4. Prompt Evaluations: Evaluation methodology
5. Tool Use: Complete tool calling implementation

Each notebook runs locally. Code along as you learn. Compared to Skilljar's video courses, these suit developers who prefer learning by doing. The notebooks use lower-cost models (Claude 3 Haiku), keeping API costs minimal.

Anthropic Cookbook

The Anthropic Cookbook collects advanced implementation examples: PDF processing, embeddings, citations, and more. Best for developers who have completed foundational courses and want reference implementations for specific features.

Claude Tutorials

claude.com/resources/tutorials offers video and written guides covering Claude Code remote control, Cowork plugin development, model selection, and other topics.

Where Should You Start?

You do not need all 13. Pick the path closest to your goal.

"I have never used Claude and want to try it" -> Claude 101 -> AI Fluency: Framework & Foundations

"I am a developer building Claude into a product" -> Building with the Claude API -> your cloud platform course

"I already use Claude Code and want to go deeper" -> Claude Code in Action -> Introduction to Agent Skills -> Plus our hands-on First Hour with Claude Code tutorial

"I want to connect Claude to external tools" -> Introduction to MCP -> MCP: Advanced Topics

"I am a teacher wanting to use AI in my classroom" -> AI Fluency: Framework & Foundations -> AI Fluency for Educators -> Teaching AI Fluency

All courses on Anthropic Academy. Free. No time limits. Your pace.