DeepLearning.AI
BeginnerAI for Everyone ↗
Andrew Ng's non-technical tour of what AI can and can't do, and how to bring it into a team.
- Length
- About 7 hours over 4 weeks
- Licence
- Free to take; certificate is paid. All rights reserved.
25 courses · 17 providers · all free
The best free AI courses.
Our picks from universities, AI labs and big tech, for after you finish our learning paths. Each link goes to the provider's own site. We don't host or copy any of these courses. Some need a free account, and certificates can cost extra.
Start here
No coding needed. Good if you're new to AI or want the big picture.
DeepLearning.AI
BeginnerGenerative AI for Everyone ↗
How generative AI works, common uses, and its limits, with no coding.
- Length
- About 5 hours over 3 weeks
- Licence
- Free to take. All rights reserved.
University of Helsinki and MinnaLearn
BeginnerElements of AI ↗
A gentle, text-based course on what AI is, with short exercises and no maths required.
- Length
- Self-paced
- Licence
- Free. All rights reserved.
Google Cloud Skills Boost
BeginnerIntroduction to Generative AI (learning path) ↗
Short videos and quizzes on generative AI, LLMs and responsible AI from Google Cloud.
- Length
- 4 short activities
- Licence
- Free. All rights reserved.
NVIDIA Deep Learning Institute
BeginnerGenerative AI Explained ↗
A self-paced overview of generative AI concepts, uses and challenges.
- Length
- 2 hours
- Licence
- Free with an NVIDIA account. All rights reserved.
IBM SkillsBuild
BeginnerArtificial intelligence learning catalogue ↗
IBM's free catalogue of short AI courses with digital credentials.
- Length
- Varies by course
- Licence
- Free. All rights reserved.
Microsoft
BeginnerAI for Beginners ↗
A full curriculum from symbolic AI to neural networks, with notebooks you can run.
- Length
- 12 weeks, 24 lessons
- Licence
- MIT licence
Machine learning
The foundations: how models learn from data. Some Python and maths helps.
Machine Learning Crash Course ↗
Fast, practical ML basics with interactive visuals and exercises.
- Length
- 12 self-contained modules
- Licence
- Text CC BY 4.0; code Apache 2.0
Microsoft
BeginnerML for Beginners ↗
Classic machine learning with Scikit-learn, lesson by lesson, with quizzes.
- Length
- 12 weeks, 26 lessons
- Licence
- MIT licence
MIT OpenCourseWare
Intermediate6.036 Introduction to Machine Learning (Fall 2020) ↗
MIT's undergraduate ML course: lecture videos, notes and exercises.
- Length
- One semester
- Licence
- CC BY-NC-SA 4.0
MIT OpenCourseWare
Intermediate6.034 Artificial Intelligence (Fall 2010) ↗
Patrick Winston's classic lectures on search, reasoning and learning.
- Length
- One semester of lectures
- Licence
- CC BY-NC-SA 4.0
Harvard University
IntermediateCS50's Introduction to Artificial Intelligence with Python ↗
Search, logic, probability, optimisation and neural networks, built up through Python projects.
- Length
- 7 weeks
- Licence
- CC BY-NC-SA 4.0
Stanford University
AdvancedCS229 Machine Learning ↗
The mathematical foundations of machine learning, from Stanford's main ML course.
- Length
- One quarter (2018 lectures on YouTube)
- Licence
- Free to watch. Not openly licensed.
Deep learning and LLMs
Neural networks, transformers and large language models, mostly hands-on in code.
fast.ai
IntermediatePractical Deep Learning for Coders ↗
Top-down, code-first deep learning: train real models from the first lesson.
- Length
- 9 lessons of about 90 minutes
- Licence
- Free. Book code GPL-3.0; text not openly licensed.
MIT
Intermediate6.S191 Introduction to Deep Learning ↗
MIT's fast-moving bootcamp on deep learning, with labs in Python.
- Length
- One-week bootcamp; lectures posted online
- Licence
- Labs MIT licence
Hugging Face
IntermediateLLM Course ↗
Transformers, tokenisers, fine-tuning and sharing models with the Hugging Face libraries.
- Length
- 12 chapters
- Licence
- Apache 2.0
Hugging Face
AdvancedDiffusion Models Course ↗
The theory behind diffusion image models, and how to train and fine-tune them.
- Length
- Self-paced units
- Licence
- Apache 2.0
Stanford University
AdvancedCS224N Natural Language Processing with Deep Learning ↗
Word vectors to transformers and LLMs, from Stanford's NLP course.
- Length
- One quarter (2023 lectures on YouTube)
- Licence
- Free to watch. Not openly licensed.
Andrej Karpathy
AdvancedNeural Networks: Zero to Hero ↗
Build neural networks and a GPT from scratch in code, one video at a time.
- Length
- Video series
- Licence
- Free to watch. Code MIT licence.
Building with AI
Prompting, apps and agents, using today's models through their APIs.
DeepLearning.AI with OpenAI
BeginnerChatGPT Prompt Engineering for Developers ↗
Prompting patterns for developers calling an LLM from code.
- Length
- 9 lessons, about 1.5 hours
- Licence
- Free. All rights reserved.
Microsoft
BeginnerGenerative AI for Beginners ↗
Build generative AI apps step by step: prompting, RAG, agents and more.
- Length
- 21 lessons
- Licence
- MIT licence
Microsoft
BeginnerAI Agents for Beginners ↗
Agent design patterns, tool use, planning and multi-agent systems, with code samples.
- Length
- 18 lessons
- Licence
- MIT licence
Hugging Face
IntermediateAI Agents Course ↗
Build and deploy agents with popular frameworks, ending in a graded project.
- Length
- 4 units plus bonus units
- Licence
- Apache 2.0; free certificate
Anthropic
BeginnerClaude Academy ↗
Courses on AI fluency and building with Claude, from Anthropic.
- Length
- Several short courses
- Licence
- Free. All rights reserved.
OpenAI
BeginnerOpenAI Academy ↗
OpenAI's free learning hub for using ChatGPT and building with its API.
- Length
- Videos, workshops and guides
- Licence
- Free. All rights reserved.
Links, lengths and licences checked October 20, 2026. Licences are shown for information only; read each provider's terms before reusing their material. No provider paid to be listed. Want the papers and docs behind our own lessons? See Learn sources.