--- title: "Stanford's free AI courses: what is actually free, and which ones suit a beginner" url: https://aiguideson.com/blog/stanford-free-ai-courses-which-suit-beginners/ summary: "Stanford puts some AI course material online free, mostly lecture videos, slides and code. The certificates, credit and marking cost money or are for enrolled students. The well-known courses expect programming and university maths, so if you do not code, start with a beginner course and come back to Stanford later." published: 2026-10-27 updated: 2026-10-27 author: "Jack Stovell" publisher: "Adapt Progress Evolve Limited" language: en-GB --- # Stanford's free AI courses: what is actually free, and which ones suit a beginner *By Jack Stovell. Published 27 October 2026.* Stanford puts some AI course material online free, mostly lecture videos, slides and code. The certificates, credit and marking cost money or are for enrolled students. The well-known courses expect programming and university maths, so if you do not code, start with a beginner course and come back to Stanford later. Short answer: what you usually get free is the teaching material, meaning recorded lectures, slides, notes or code. The marking, the support, the credit and any certificate cost money or are only for enrolled students. The best-known courses also expect you to program and to have university maths before you start. AI Guides On is not affiliated with Stanford, and Stanford has not endorsed this post. Everything below comes from Stanford's own pages, read on 9 October 2026. ## Checking the "full Stanford AI degree for free" claim against Stanford's own pages Posts on social media claim that Stanford's AI degree is now free, often with a large dollar figure attached. None of the Stanford pages checked for this post offers a free degree. What they offer is material from individual courses. A degree needs admission, credit and graded work, and the course pages make clear that those belong to enrolled students. The natural language processing course, CS224N, sets this out most clearly. Its page describes three ways in: - **Free:** complete lecture videos from the 2024 course on YouTube, plus slides and assignments posted online that the course says anyone is welcome to use. - **XCS224N**, through the Stanford Artificial Intelligence Professional Program: described as "medium fee, community TAs and certificate". - **CS224N through Stanford Online:** described as "high cost, limited enrollment, gives Stanford credit". So the lectures are free, and the certificate and the credit are paid for. The page also says videos from the current term are for enrolled students only. The Stanford Online free-courses page is the obvious place to check whether its free courses come with a certificate. When we checked, the list only loaded with JavaScript, and the text we could read said nothing about certificates, credit or degrees. This post therefore makes no claim either way. Before you assume there is a certificate, open the course you want and look for the word "certificate" on its own page. ## What the well-known Stanford AI courses assume you already know Each course lists its prerequisites on its own page. In plain terms: - **CS229, Machine Learning** (Summer 2026 page): you can write "a reasonably non-trivial computer program in Python/NumPy", and you know probability, multivariable calculus and linear algebra. - **CS221, Artificial Intelligence: Principles and Techniques** (Autumn 2026 page): programming, discrete maths, probability and linear algebra are all required, and algorithms is recommended. The page warns that the course "is fast-paced and covers a lot of ground". - **CS224N, Natural Language Processing with Deep Learning**: you are proficient in Python, because the assignments use NumPy and PyTorch. You also need college calculus and linear algebra, and you should be comfortable with multivariable derivatives and matrix notation. These courses are written for Stanford students who have already taken the earlier courses they name. If the list looks unfamiliar, these courses come later in your learning. You can still understand AI well without them. ## Free to watch, free to audit and free with a certificate are different things These phrases sound alike, so it helps to separate them. **Free to watch** means the lectures are posted and you can press play. Nobody marks your work or checks that you understood. The CS224N YouTube videos are this kind. **Free materials** means slides, notes or code you can download. CS221 puts its slides and code on GitHub, but its lecture recordings are on Stanford's Canvas, which needs a Stanford login. The CS229 page says its course documents are only shared with Stanford University affiliates. **Free to audit** is a phrase some course platforms use for following along without the graded work or a certificate. None of the three course pages checked here uses it, so read the terms on whichever platform hosts the course you want. **With a certificate or credit** is the version people usually hope for. On CS224N, both of those are paid options. Our post on [what to check before trusting a free AI course with a certificate](/blog/free-ai-courses-with-certificate-what-to-check/) has a checklist that works for any provider. ## Stanford material a non-coder can realistically use **CRAFT.** Stanford's CRAFT site describes its resources as free and "for non-profit use by classroom educators". They were designed with practising high school teachers from across the United States, and they range from short activities to full lessons. They are aimed at teachers, so expect classroom material rather than a course for adults studying alone. They still give you a plain-language view of how AI is taught. **Course descriptions and topic lists.** The top of each course page is readable without any maths. CS229 lists topics such as supervised learning, neural networks and reinforcement learning. CS224N describes how large language models such as ChatGPT fit into the field. Reading these gives you the vocabulary. **A lecture or two, for the feel of it.** You can watch an early CS224N lecture on YouTube to hear how researchers talk about language models. Expect the maths to go over your head. **Skip for now:** the assignments, the problem sets and anything that asks you to set up Python. Without the prerequisites, that is where these courses stop making sense. ## A sensible order if you are starting from zero 1. **Start with a course that assumes nothing.** [AI for beginners: start here](/courses/ai-for-beginners-start-here/) is 5 lessons, about 45 minutes in total, and you can read it without an account. It opens with [what AI chat tools are, and what they get wrong](/courses/ai-for-beginners-start-here/what-ai-chat-tools-are/). 2. **Learn to ask well.** [Prompting for beginners](/courses/prompting-for-beginners/) shows you how to get useful answers from the tools you will actually use. 3. **Then go back to Stanford.** Reread a course page and see how much more makes sense. If you want to take one properly, the prerequisites list tells you what to learn first: basic Python, then probability and linear algebra. If you finish one of our courses, you can get a certificate of completion. It records that you finished the lessons. It is not a qualification, and it has nothing to do with Stanford. The [certificates page](/certificates/) explains what it shows. ## What to do this week 1. Read the first lesson of AI for beginners: start here. 2. Open the CS224N page, find the prerequisites list and write down any words you do not know. 3. Ask an AI chat tool to explain one of those words in plain English. Then check its answer using the steps in [check what it tells you](/courses/ai-for-beginners-start-here/check-what-it-tells-you/). ## Sources - [CS229: Machine Learning, Stanford course page (Summer 2026), fetched 9 October 2026](https://cs229.stanford.edu/) - [CS221: Artificial Intelligence: Principles and Techniques, Stanford course page (Autumn 2026), fetched 9 October 2026](https://stanford-cs221.github.io/autumn2026/) - [Stanford CS224N: Natural Language Processing with Deep Learning, course page, fetched 9 October 2026](https://web.stanford.edu/class/cs224n/) - [CRAFT, Stanford: home page, fetched 9 October 2026](https://craft.stanford.edu/) - [CRAFT, Stanford: About, fetched 9 October 2026](https://craft.stanford.edu/about/) ## Questions people ask ### Are Stanford's AI courses really free? Some of the material is. For example, the CS224N page links to complete lecture videos on YouTube and lets anyone use its slides and assignments, and CS221 puts its slides and code on GitHub. The certificate and the Stanford credit are paid options, and some recordings are for enrolled students only. ### Do I need to code or know maths to take a Stanford AI course? For the well-known ones, yes. CS229, CS221 and CS224N all list programming (Python for CS229 and CS224N) and university-level maths, such as linear algebra, as prerequisites on their own pages. Their course descriptions and topic lists are readable without either. ### Do free Stanford courses give you a certificate or a degree? None of the Stanford pages checked for this post offers a free degree. On CS224N, the certificate comes with the paid XCS224N version and the credit comes with the paid Stanford Online version. The Stanford Online free-courses page did not say anything we could read about certificates, so check each course's own page before you assume one. ### Which Stanford AI course is best for a complete beginner? Going by their stated prerequisites, none of CS229, CS221 or CS224N is written for a complete beginner. Stanford's CRAFT resources use plain language, but they are made for high school teachers to use in class, not as a course for adults. ### Where do I start if I want to learn AI but do not code? Start with a course that assumes nothing, such as AI for beginners: start here. It is free to read without an account. Then move on to Prompting for beginners, and come back to the Stanford course pages once you have the basics.