Learning AI in 2026 can be overwhelming, given the constant influx of new courses, tutorials, repositories, and frameworks. The challenge is no longer finding resources, but determining what to learn first and how to apply that knowledge. To address this, I have curated 25 free AI resources organized into a practical learning path, covering machine learning fundamentals, LLMs, RAG, AI agents, and generative AI.
25 Free Resources to Learn AI, LLMs and AI Agents
Below are my top recommended free resources for learning AI, LLMs, and AI agents from the ground up.
Stage 1: Build Your AI/ML Foundation
Begin here if you are new to AI or transitioning from a data analysis background:
- Google Machine Learning Crash Course
- Kaggle Learn
- fast.ai Practical Deep Learning
- Andrej Karpathy’s Neural Networks: Zero to Hero
- 10 Beginner Machine Learning Projects
Stage 2: Learn LLMs and Generative AI
Once you are confident with Python and machine learning fundamentals, proceed to LLMs:
- Hugging Face LLM Course
- Hugging Face Learn
- Google’s Introduction to Generative AI
- Andrej Karpathy’s LLM Videos
- 20 Real-World LLM Projects
Stage 3: Master RAG and Prompt Engineering
Next, explore RAG and prompt engineering using the following resources:
- LangChain Documentation
- LlamaIndex Documentation
- OpenAI Prompt Engineering Guide
- Advanced Retrieval for AI with Chroma (Short Course)
- 10 Real-World RAG Projects
Preparing for AI/ML Interviews?
If you are using this roadmap to become job-ready, I recommend preparing for interviews as you progress. Cracking Your First AI/ML Interview is a valuable resource to help translate your learning into interview-ready skills.
Stage 4: Learn AI Agents
At this stage, LLM applications become significantly more advanced and engaging:
- Hugging Face AI Agents Course
- LangGraph Documentation
- CrewAI Documentation
- Microsoft’s AI Agents for Beginners
- 20 AI Agent Projects That Solve Real Business Problems
Stage 5: Build Production-Ready AI Systems
After successfully running an agent locally, continue with these resources to advance your skills:
- Docker Documentation
- GitHub Actions Documentation
- Hugging Face Agents Evaluation
- Hugging Face Context Course
- 50 Real-World AI Projects
The Takeaway
It is not necessary to complete all 25 resources. Instead, progress through each stage in order and create a project after every major milestone.
Learn machine learning, then build an ML project. Study LLMs, then develop a local LLM application. Explore RAG, then create a document intelligence system. Understand agents, then implement a tool-using or multi-agent system. Learn deployment, then move one of your projects into production.
The most common mistake learners make is accumulating courses rather than building projects. You do not need numerous certificates; instead, focus on a few projects that you understand thoroughly and can explain from architecture through implementation and failure handling.
Thank you for reading this overview of 25 free resources to learn AI, LLMs, and AI agents from scratch. For additional tips on AI and machine learning, feel free to follow me on Instagram.
