huggingface/agents-course

★ 28,714MDXAudience · developerComplexity · 2/5ActiveLicenseSetup · easy

Mindmap

mindmap
  root((repo))
    What it does
      Teaches AI agents
      Hands-on frameworks
      Four main units
    Tech stack
      Python
      LLMs
      smolagents
      LlamaIndex
      LangGraph
    Use cases
      Learn agent building
      Build web search agents
      Create API integrations
      Production workflows
    Learning path
      Basics first
      Framework deep dives
      Bonus topics
      Certification project
    Audience
      Python developers
      AI enthusiasts
      Non-technical curious

mindmap root((repo)) What it does Teaches AI agents Hands-on frameworks Four main units Tech stack Python LLMs smolagents LlamaIndex LangGraph Use cases Learn agent building Build web search agents Create API integrations Production workflows Learning path Basics first Framework deep dives Bonus topics Certification project Audience Python developers AI enthusiasts Non-technical curious

Things people build with this

USE CASE 1

Learn how to build AI agents that reason and take actions using language models.

USE CASE 2

Build agents that search the web, call APIs, or write and run code automatically.

USE CASE 3

Get hands-on experience with smolagents, LlamaIndex, and LangGraph frameworks.

USE CASE 4

Create production-ready agent workflows and benchmark your own agent for certification.

Tech stack

PythonLLMssmolagentsLlamaIndexLangGraphMDX

Getting it running

Difficulty · easy Time to first run · 5min

Use freely for any purpose including commercial. Keep the notice and disclose changes to the patent grant.

In plain English

This repository contains the Hugging Face Agents Course, a free, structured online curriculum that teaches you how to build AI agents from the ground up. An AI agent is a program that uses a large language model (LLM, the kind of AI behind tools like ChatGPT) to reason through problems and take actions, like searching the web, calling APIs, or writing and running code. The course is divided into four units. It starts with the basics of what agents are and how LLMs work, then moves into hands-on coverage of three popular frameworks for building agents: smolagents (a lightweight Hugging Face library), LlamaIndex (for building agents that work with your own data), and LangGraph (for building more production-ready, controllable agent workflows). There are also bonus units on fine-tuning models, tracing and evaluating agents, and even using agents in games. The final unit involves building and benchmarking your own agent for certification. It requires basic Python knowledge and some familiarity with LLMs. The course material is written in MDX (a format that mixes text and interactive components) and is freely available at the Hugging Face learning platform. You would use this course if you are a developer or curious non-technical person wanting to understand how AI agents work and how to build them using today's most common tools, without starting from scratch.

Copy-paste prompts

Prompt 1

I want to build an AI agent that can search the web and summarize results. Which framework from the Hugging Face Agents Course should I start with?

Prompt 2

Show me how to use smolagents to create an agent that calls external APIs and processes the responses.

Prompt 3

I need to build a production-ready agent workflow. Walk me through the LangGraph approach covered in the Hugging Face Agents Course.

Prompt 4

How do I fine-tune a language model for my agent based on the bonus units in the Hugging Face Agents Course?

Prompt 5

Help me set up a basic agent using the smolagents framework following the Hugging Face Agents Course structure.

Open on GitHub → Explain another repo

Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.