Free courses, in order¶
Last reviewed: October 2026
There are hundreds of free AI courses. You need a few, in the right order. Pick the stage you're at and take one course from it before moving on. Every link below is free (some Coursera courses are free to audit; you pay only for a certificate).
Learning for an interview?
Courses build understanding; interviews test whether you can explain and defend it. After each stage, open the matching topic in this library and practise its questions in the OfferReady app.
Stage 1 · First concepts (no code)¶
| Course | Who | Why take it |
|---|---|---|
| Elements of AI | University of Helsinki | What AI is and isn't, in plain language. A few evenings. |
| AI for Everyone | DeepLearning.AI | How AI projects work in a company. Free to audit. |
| Neural networks | 3Blue1Brown (videos) | The clearest visual explanation of how networks and transformers learn. |
Stage 2 · Python and data¶
| Course | Who | Why take it |
|---|---|---|
| Python | Kaggle | Just enough Python, in your browser. |
| All Kaggle Learn courses | Kaggle | Pandas, data cleaning and visualisation, each a few hours. |
| CS50's Intro to AI with Python | Harvard | Search, logic, ML and neural nets, with real projects. |
Stage 3 · Machine learning foundations¶
| Course | Who | Why take it |
|---|---|---|
| Intro to Machine Learning | Kaggle | Train and validate your first models. |
| Machine Learning Crash Course | The core ideas with exercises. | |
| Practical Deep Learning | fast.ai | Build working models first, theory second. |
| CS229 Machine Learning | Stanford | The maths, for when you want depth. |
Stage 4 · LLMs and generative AI¶
| Course | Who | Why take it |
|---|---|---|
| LLM Course | Hugging Face | Transformers, tokenizers and fine-tuning, chapter by chapter. |
| Learn Prompting | Learn Prompting | Prompting techniques from basic to advanced. |
| Short courses | DeepLearning.AI | One-hour courses on RAG, evals and agents, often with the tool makers. |
| Anthropic Academy | Anthropic | Claude, the API, MCP and Claude Code. |
| Vector database learning | Zilliz | Embeddings, indexes and vector search explained. |
Stage 5 · Agents and production¶
| Course | Who | Why take it |
|---|---|---|
| AI Agents Course | Hugging Face | Build agents with tools, memory and evaluation. |
| Introduction to LangGraph | LangChain Academy | Agents as graphs: state, memory, human-in-the-loop. |
| MLflow docs | MLflow | Tracking experiments, models and LLM traces. |
| Deep RL Course | Hugging Face | Reinforcement learning, if your role needs it. |
Then: prove it¶
Courses alone don't get offers. For each stage, build one small thing (a notebook, a RAG demo, an agent with two tools), write down one decision you made and why, and practise explaining it. Start with Agentic AI in 7 diagrams to learn to draw and explain the core designs.