What is MIT OpenCourseWare and how does it work?
OpenCourseWare, or OCW, is MIT's archive of the materials from its own campus classes: recorded lectures, notes, problem sets and exams, with solutions where the class provided them. There is no enrolment, no start date and no account, and the content is licensed Creative Commons BY-NC-SA, so you may download, share and adapt it for non-commercial use with credit. Courses marked SC are Scholar editions, arranged for independent study.
Is it free, and does it give certificates?
OCW is free in full and never issues certificates or grades. If a credential matters, the same subject often exists as an MITx course where a certificate can be bought; the sections below point to those pairs.
OpenCourseWare vs MITx: the same subject, two different products
MITx Online courses, now presented on learn.mit.edu, are built for online learners rather than recorded from a lecture hall. Every MITx course listed here has a free audit track with the full course and its materials; the paid certificate track adds graded assignments and exams and an MIT verified certificate of completion. Some are instructor-paced with scheduled runs, others self-paced; the course page says which.
A third channel, the Open Learning Library, hosts selected MITx and OCW courses on an interactive platform: exercises give instant feedback and a free account tracks progress. There are no fees and no certificates there; MIT points anyone who wants a credential to the MITx version.
How to use MIT OpenCourseWare
Every OCW course works the same way, and none of it needs a login.
- Open the course on ocw.mit.edu and read the syllabus; it lists prerequisites, and course numbers signal the order (6.006 before 6.046J).
- Work through the lecture videos in order with the notes and worked examples.
- Attempt the problem sets and exams before opening the solutions.
- Use the download option for an offline copy; the licence allows it.
- For auto-graded practice, find the course in the Open Learning Library; for a certificate, find the MITx version.
MIT free online courses with certificates: what the paid track costs
There is no free MIT certificate. Every certificate here comes from an MITx course and is paid, with the price on the course page. The cheapest are $49 for Science of Learning and Memory and $49 for Entrepreneurship 101: Who is Your Customer. Introductory programming sits at $149 and calculus at $100.
MicroMasters courses cost more: $300 for statistics and $300 for machine learning. Financial Accounting is $549 and Foundations of Modern Finance I is $549. Microeconomics uses a sliding scale with financial aid. Other universities are on the courses with certificates page.
What you can study on OCW and MITx
Each cluster names the OCW option first, then any graded alternative.
Linear algebra with Gilbert Strang
Linear Algebra is the Scholar edition of Gilbert Strang's 18.06, in three units from elimination through eigenvalues to positive definite matrices; it also sits in the Open Learning Library with interactive exercises. Strang's Matrix Methods in Data Analysis, Signal Processing, and Machine Learning applies that material to deep learning.
Calculus, and why multivariable calculus is missing
Single Variable Calculus covers differentiation, integration and a short treatment of series. The graded MITx equivalents are Calculus 1A: Differentiation and Calculus 1B: Integration. Calculus 1A runs fourteen weeks at 8 to 10 hours a week, and Calculus 1B fifteen weeks. The certificate is $100 for 1A and $100 for 1B. No multivariable calculus course is listed here. More are on the mathematics hub.
AI, machine learning and deep learning
Artificial Intelligence is Patrick Winston's classical course on search, constraints and symbolic learning, not a deep-learning toolkit. Introduction to Deep Learning is the short January course with TensorFlow labs. For graded practice without a certificate, Introduction to Machine Learning runs thirteen weeks in the Open Learning Library. The certificate route is Machine Learning with Python. It runs fifteen weeks at 10 to 14 hours a week, with a $300 certificate.
Computer science, Python and algorithms
Start with Introduction to Computer Science and Programming in Python, twelve OCW lectures for beginners, then Introduction to Computational Thinking and Data Science. The MITx versions, 6.00.1x and 6.00.2x, are instructor-paced. 6.00.1x runs nine weeks at 14 to 16 hours a week, and 6.00.2x nine weeks at 11 to 13 hours a week. The certificate is $149 for 6.00.1x and $149 for 6.00.2x.
For algorithms, Introduction to Algorithms has twenty lectures and expects you to know a language; Design and Analysis of Algorithms follows it, and Mathematics for Computer Science supplies the proofs. Other universities are on the computer science hub.
Statistics and probability
Statistics for Applications is the OCW course on mathematical statistics. On MITx, Probability - The Science of Uncertainty and Data and Fundamentals of Statistics are the MicroMasters pair. Probability runs sixteen weeks at 10 to 14 hours a week with a $300 certificate. Fundamentals of Statistics runs seventeen weeks at 10 to 14 hours a week with a $300 certificate.
Economics, finance and physics
Principles of Microeconomics gives you Jonathan Gruber's lectures with self-check quizzes; his MITx Microeconomics adds graded work, and AP Microeconomics is the gentle self-paced version ($69). In finance, Finance Theory I is OCW and Adaptive Markets is free in the Open Learning Library. For physics, Classical Mechanics is the only course listed and assumes basic calculus.
Is OCW good enough to learn from on its own?
You get the same material an MIT student had, including exams and usually solutions, and you can download it. You do not get feedback, deadlines or proof of completion. If you can keep your own schedule and mark your own work, that is enough; if you need structure or a credential, audit the MITx course instead.