Free Online Computer Science Courses

Free online computer science courses with certificates: which ones qualify

The 59 courses on this page fall into three groups by what you get at the end: a free certificate, a paid one, or none. The provider decides, not the difficulty: a graduate-level MIT course carries no credential while a short IBM course carries a free one.

Free certificate on completion

The CS50 courses issue a free CS50 Certificate at a unique URL once you score at least 70% on every problem set or project. CS50x is the main one, and CS50 for Lawyers, CS50's Introduction to Programming with Scratch, CS50 Web Programming and CS50 AI work the same way. A free edX account is needed to submit work for feedback.

freeCodeCamp publishes each certificate at a public URL on freecodecamp.org after you finish the required projects and pass an exam that is free to take; Responsive Web Design, JavaScript, Back-End Development and APIs and Foundational C# with Microsoft are listed here. IBM's Cognitive Class courses such as SQL and Relational Databases 101 (5 hours) and Docker Essentials (3 hours) offer a certificate at no cost on completion. AWS Knowledge Badges, for example the Compute readiness path at 16 hours 20 minutes, are issued through Credly after an assessment passed at 80% or higher.

Free to audit, certificate paid

University courses on edX and MITx Online open the videos and readings without charge but put graded work and the certificate on a paid track, and edX audit access lapses after the estimated course length. Stanford's Computer Science 101 lists its verified certificate at $249, Algorithms: Design and Analysis, Part 1 at $189, the Linux Foundation's Introduction to Linux at $209, BerkeleyX's Introduction to Python at $200 and MITx's 6.00.1x at $149. CS50x also runs on edX, where the verified certificate costs $219; the free certificate on Harvard's own site covers the same work.

Edinburgh's Code Yourself! on Coursera is the exception in this group: it is free in full rather than free to audit, and its shareable certificate is a paid option with no price stated.

No certificate at all

MIT OpenCourseWare, Stanford Engineering Everywhere and Khan Academy publish complete courses with no credential, and Princeton's courses on Coursera state that no certificates, credentials or reports are awarded under university policy. Microsoft Learn paths such as GitHub Foundations end with a profile trophy, not a certificate.

Harvard or MIT: how their free computer science courses differ

Harvard's free computer science course is CS50x, which runs 11 weeks at 10 to 20 hours a week, moves from C to Python, SQL and web development, and ends with a final project. The Harvard hub lists the rest of the CS50 family.

MIT publishes its material two ways. OpenCourseWare gives you the recorded lectures, problem sets and exams of 6.0001, 6.006 Introduction to Algorithms and 6.042J Mathematics for Computer Science with no account, no deadlines and no certificate. MITx Online runs graded versions of the introductory sequence, 6.00.1x (9 weeks at 14 to 16 hours a week) and 6.00.2x, free to audit with a paid certificate. See the MIT hub and whether OpenCourseWare gives certificates.

Which introduction to computer science online course to start with

The starting point depends on whether you have written code before and which language you want to meet first.

Never written code

CS50's Introduction to Programming with Scratch (9 weeks) and Edinburgh's Code Yourself! use block-based Scratch to teach functions, loops and variables before any typing. Stanford's Computer Science 101 (6 weeks at 4 to 6 hours a week) runs small pieces of code in the browser and covers hardware, images, the internet and security. Khan Academy's AP Computer Science Principles follows the AP CSP curriculum for students preparing for that exam.

Some code, want the full introduction

CS50x is the standard answer. Alternatives by language: Princeton's Programming with a Purpose teaches Java from Sedgewick and Wayne's textbook; BerkeleyX's Introduction to Python (7 weeks) covers functions, iteration and higher-order functions; Stanford's CS106A is the recorded campus course in Java with assignments and exams.

After the introduction

The intermediate courses split by thread, and most assume you can already program.

For discrete mathematics, MIT's 6.042J covers proofs, graph theory, counting and probability and names single-variable calculus as its only prerequisite.

• Algorithms: MIT's 6.006, Princeton's Algorithms, Part I (5 weeks) and Part II in Java, and Stanford's Algorithms: Design and Analysis, Part 1 (6 weeks).
• Systems: Princeton's Computer Architecture, Stanford's CS107 Programming Paradigms and MIT's graduate Computer Systems Security.
• Software engineering: MIT's Software Construction in Java (14 weeks) on testing, specifications and abstract data types.
• Artificial intelligence and machine learning: CS50 AI (7 weeks at 10 to 30 hours a week), MIT's Introduction to Machine Learning (13 weeks) and Stanford's CS229.

Do you need mathematics or programming before you start

Not for the introductions: CS50x, the Scratch course, Code Yourself! and Computer Science 101 state no prerequisites, and Stanford's CS106A says it needs no programming background but considerable dedication. The intermediate courses are explicit about what they assume. CS50 AI asks for CS50x or a year of Python; Stanford's algorithms course says it is not an introduction to programming; 6.046J assumes 6.006; MIT's machine learning course recommends Python, calculus and linear algebra; Mining Massive Datasets expects prior courses in data structures, algorithms, databases, linear algebra, multivariable calculus and statistics.

Related pages on this site

For the practical side of the same material, the programming and Python pages list courses by language rather than by theory, and the IT page covers Linux, cloud and networking. Mathematics has the calculus and linear algebra the machine learning courses assume, and cyber security continues from MIT's systems security course. Free credentials across subjects are collected at free certifications.