Guide · Open source
How to use ai-agents-for-beginners: clone, set up and work through the lessons
Updated · written with AI from the project's own documentation and checked against our catalogue data
ai-agents-for-beginners is a free, MIT-licensed course from Microsoft, not a software product, that teaches the basics of building AI agents. It is aimed at developers new to agents, students, and people who already know basic generative AI. By the end of this guide you will know how to clone the repository, what you need before you start, and how to pick and work through a lesson.
AI Agents for Beginnersmicrosoft/ai-agents-for-beginners · ★ 78.7k · MIT Install options & downloads Quick start
- 1Check prerequisites. Make sure you have an Azure account for Microsoft Foundry and Python installed.
- 2Clone with sparse checkout. Run the git clone command with --filter=blob:none --sparse to fetch the repository without translations.
- 3Set the sparse checkout. Enter the ai-agents-for-beginners folder and run git sparse-checkout set with the exclusions for translations and translated_images.
- 4Read the course setup. Open the Course Setup page to configure the code samples.
- 5Pick a lesson. Choose any lesson from the lessons table, since each covers its own topic.
- 6Run the code sample. Open the matching sample in the code_samples folder and work through it.
What you need before you start
The code examples use Microsoft Agent Framework with Microsoft Foundry Agent Service V2. The README lists Microsoft Foundry as requiring an Azure account, so plan on having one before you run the main path.
The README names Python as a prerequisite but does not state a version, so check the course setup page for that detail.
- An Azure account for Microsoft Foundry (required by the README)
- Python (version not stated in the README excerpt)
- Basic familiarity with generative AI, which the course assumes
Clone the repository with sparse checkout
The repository includes more than 50 translations, which makes the full download large. The README recommends a sparse checkout that skips the translations folders. This gives you everything needed to complete the course with a faster download.
On Bash, macOS or Linux, clone and then set the sparse checkout from inside the project folder. The README also gives a CMD version for Windows, which uses double quotes instead of single quotes.
git clone --filter=blob:none --sparse https://github.com/microsoft/ai-agents-for-beginners.git
cd ai-agents-for-beginners && git sparse-checkout set --no-cone '/*' '!translations' '!translated_images'Choose a lesson and read its guide
Each lesson covers its own topic, so you can start with any lesson. Each one includes a written guide in its README, a short video, and Python code samples. Lesson guides are linked from the lessons table in the main README.
The code examples are in the code_samples folder. If you plan to change them, the README suggests forking the repository to create your own copy.
- Read the lesson README for the concept and walkthrough
- Watch the short video linked in the lessons table
- Run the matching Python code sample from the code_samples folder
Get help and follow the course setup
The course setup page, linked from the README as Course Setup, explains how to configure the code to run. The README also says some samples support alternative OpenAI-compatible providers such as MiniMax, and points to the setup page for configuration details.
If you get stuck, the README directs questions to a dedicated channel in the Microsoft Foundry Discord.
- Start with the Course Setup page for configuration
- Use the Microsoft Foundry Discord for questions
What people use it for
- Learning how an agent is structured by working through a lesson's Python code sample
- Building a first agent on Microsoft Agent Framework with Microsoft Foundry Agent Service
- Studying agentic AI concepts as a student using guided lessons and videos
- Moving from basic generative AI knowledge to agent development
Frequently asked questions
Is ai-agents-for-beginners free?
The repository is released under the MIT licence, as shown in the README. The main path, however, uses Microsoft Foundry, which the README says requires an Azure account.
Do I need an Azure account to use ai-agents-for-beginners?
The README says the main path requires Microsoft Foundry and marks it as Azure Account Required. Without an Azure account, the main code path is not available to you.
What do I need to install to run the code?
You need Python and an Azure account for Microsoft Foundry. The README does not state a Python version, so check the Course Setup page for configuration details.
Why is the git clone so large, and how do I make it smaller?
The repository includes more than 50 language translations, which increases the download size. Use the sparse checkout commands from the README to skip the translations and translated_images folders.
Is this a ready-made agent I can use without coding?
No. The course teaches you how to build AI agents and is built around Python code samples, so it is not a finished agent product.
How does it compare to LangChain, AutoGen or Semantic Kernel?
This is a course rather than a framework, and its examples centre on Microsoft Agent Framework and Microsoft Foundry. If you want a framework-neutral comparison, the README does not provide one, so look at projects such as LangChain, AutoGen and Semantic Kernel directly.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Choose this course if you are a developer, student or experienced generative AI user who wants guided lessons with runnable Python code and you have or can get an Azure account. Pick something else if you want a ready-made agent product, cannot use Azure, or need a framework-neutral comparison.
Install options & downloads