Official videos
Every video on this page is published by Anthropic on its own YouTube channel. Nothing here is third-party courseware. They are arranged by the exam they serve, in the order worth watching.
On the website each video plays inline. On GitHub the thumbnails link out to YouTube.
Tip
Video is a supplement, not a substitute. The 21 official courses are the curriculum and the exam guides define what is tested. Use these when a topic has not landed from reading, and when you want to hear how Anthropic's own engineers describe a decision.
AI Fluency, the complete course
Anthropic publishes AI Fluency: Framework & Foundations as a full video course, and it is also one of the 21 Academy courses with its own certificate. It underpins the Associate exam: the 4D framework, capabilities and limitations, and the prompting and judgment habits the scenarios assume.
The complete playlist runs about 70 minutes. The lessons that carry the most exam weight:
Lesson 2B: The 4D framework
Delegation, Description, Discernment, and Diligence, the spine of the whole course and of the Associate blueprint.
Lesson 3B: Capabilities and limitations
What Claude can and cannot do, and how to tell the difference. This is the material behind the AI Capabilities and Limitations course and the questions that ask you to verify rather than trust.
Lesson 7: Effective prompting techniques
The deep dive on prompting, and the longest lesson in the course.
Every lesson in the course, in order
| Lesson | Watch |
|---|---|
| Course trailer | Watch |
| 1. Introduction to AI Fluency | Watch |
| 2A. Why do we need AI Fluency? | Watch |
| 2B. The 4D framework | Watch |
| 3A. What is generative AI? | Watch |
| 3B. Capabilities and limitations | Watch |
| 4. A closer look at Delegation | Watch |
| 6. A closer look at Description | Watch |
| 7. Effective prompting techniques | Watch |
| 8. A closer look at Discernment | Watch |
| 10. A closer look at Diligence | Watch |
| 11. Conclusion | Watch |
Lesson numbering follows Anthropic's published playlist.
Prompting
Relevant to Associate and Developer. Prompt structure and iteration appear in the blueprint of every exam.
Prompting 101
Prompt structure, iteration, and the habits behind reliable output, recorded at the Code w/ Claude conference.
AI prompt engineering: a deep dive
A longer conversation on how prompting actually works, from the people who do it at Anthropic.
Prompting for agents
How prompting changes once the model is acting rather than answering. Relevant to the agent domains on the Developer and Architect papers.
Claude Code
Relevant to Developer and Architect Foundations, and the video counterpart to the Claude Code 101 and Claude Code in Action courses.
Mastering Claude Code in 30 minutes
A compact end-to-end tour. The fastest way to see the whole tool at once.
Claude Code best practices
Configuration, workflows, and working habits, recorded at Code w/ Claude.
Building headless automation with Claude Code
Headless mode and automation, which the Developer blueprint names explicitly.
Also worth watching: Designing Claude Code on why the tool works the way it does, Building and prototyping with Claude Code, The future of agentic coding, and a conversation on Claude Code.
The Claude API and MCP
Relevant to Developer and both Architect exams, and the video counterpart to Building with the Claude API, Introduction to Model Context Protocol, and Model Context Protocol: Advanced Topics.
Building with MCP and the Claude API
The two together, which is how the exams treat them.
MCP 201
Beyond the basics: the material the Advanced Topics course covers.
Why Anthropic built and donated MCP
The reasoning behind the protocol, which explains the design decisions the scenario questions ask you to apply.
Also worth watching: The Model Context Protocol as an introduction, and MCP at Sourcegraph for a production account.
Agents and subagents
Relevant to Developer and both Architect exams. These pair with Building effective agents, the article the agent domains are built on.
Building more effective AI agents
The workflow-versus-agent decision, which is the judgment the scenario questions test most often.
Tips for building AI agents
Practical guidance on scoping, tools, and when to stop adding autonomy.
Also worth watching: Building the future of agents with Claude, Building blocks for tomorrow's AI agents, and Lessons on AI agents from Claude Plays Pokemon.
Deploying on Bedrock and Vertex AI
Relevant to both Architect exams, and the video counterpart to the Claude with Amazon Bedrock and Claude on Google Cloud courses. The Architect Professional blueprint expects you to choose between them on evidence.
Claude in Amazon Bedrock
Claude in Google Cloud's Vertex AI
Every official playlist
Anthropic's channel is organized into playlists. These are the ones with study value, most relevant first.
| Playlist | Videos | Most relevant to |
|---|---|---|
| AI Fluency Course | 12 | Associate |
| Developers | 14 | Developer, Architect |
| Code w/ Claude Developer Conference | 20 | Developer, Architect |
| AI Fluency for nonprofits | 10 | The nonprofits course |
| Claude 3.7 Sonnet and Claude Code | 2 | Developer |
| Product | 10 | Associate |
| Research | 22 | Background, not examinable |
| How Anthropic uses Claude | 1 | Associate |
Industry playlists (Financial Services, Life Sciences) are useful context for customer conversations but are not on any blueprint.
Note
Video titles and playlist contents change as Anthropic publishes. Links were last checked on 27 July 2026, and the weekly link check catches any that break.
Facts drawn from the official Anthropic exam guides. Repository index